Global Technology

Global Technology| July 22, 2015 MORGAN STANLEY & CO. LLC July 22, 2015 Katy L. Huberty, CFA [email protected] +1 212 761-6249 GGlloobbaall TTeecchhnnoollooggyy Joseph Moore [email protected] +1 212 761-7516 Global Insight: Cloudy with a Chance of James E Faucette Revisions [email protected] +1 212 296-5771 Keith Weiss, CFA [email protected] +1 212 761-4149 Our new proprietary global cloud capex tracker highlights growth MORGAN STANLEY ASIA LIMITED+ deceleration that will weigh on server & component suppliers. We Jasmine Lu lower ests. on 11 global technology stocks; downgrade Aspeed, [email protected] +852 2239-1348 MORGAN STANLEY TAIWAN LIMITED+ QLogic, Seagate, WD & our US IT Hardware view to Cautious. Charlie Chan [email protected] +886 2 2730-1725 Chinks in cloud armor begin to show: 2015 capex growth of 20% is Grace Chen down from 37% last year. We capture $85B of capex plans for 19 of the [email protected] +886 2 2730-2890 largest contributors to data center spend, including cloud and consumer Sharon Shih Internet service providers, and account for 40%+ of web traffic and enterprise [email protected] +886 2 2730-2865 cloud spending. We estimate 20% growth in 2015 based on a combination of MORGAN STANLEY & CO. INTERNATIONAL PLC, SEOUL BRANCH+ company guidance and our estimates. This is a deceleration from strong 37% Shawn Kim growth last year and from our original 2015 capex growth forecast of 24%. [email protected] +82 2 399-4940 Reported 1Q15 capex spend represented 21% of total year investment, up MORGAN STANLEY & CO. LLC from 19% historically, pointing to a more front-end loaded year. Yet 1H15 Brian Nowak, CFA [email protected] +1 212 761-3365 estimates suggest the portion of 1H15 spending is in line with the 3-year MORGAN STANLEY MUFG SECURITIES CO., LTD.+ average. These data points make us cautious on server & storage growth Shoji Sato headed into the historically strong 3Q for cloud-related investments. [email protected] +81 3 6836-8404 More cautious on server & storage suppliers; downgrading several stocks. Cloud data center growth has provided a cushion to offset weak spend IT Hardware in traditional enterprise & consumer segments in the past 3 years. We think North America the cushion may thin NT and offer greater visibility on underlying weakness in IndustryView Cautious enterprise & commercial segments. Our June CIO survey pointed to incremental enterprise IT budget weakness, which combined with the cloud capex downtick drives several est & rating cuts across our global tech research We downgrade the US IT Hardware industry view from coverage. We lower ests for 11 global tech stocks with server exposure & In-Line to Cautious, and ratings for Aspeed (from OW to above-normal seasonal growth in 2H15. We cut ests as much as 27% (for EW), QLogic (OW to EW), Seagate (OW to UW), Western WDC) & PTs as much as 30%+ due to lower forecasts & multiple compression. Digital (OW to EW). We lower estimates for 11 global On cloud-exposed semi names, we trim our ests but maintain ratings for technology stocks with exposure to decelerating cloud capex Broadcom, Cavium, Marvell & Avago – and highlight our concern that Intel’s growth (see page 4). LT assumptions in data center are too optimistic. As a result of significant est risk, we downgrade Aspeed, QLogic, Seagate & Western Digital. We see greater upside at companies with strategic catalysts (HP), high exposure to Apple (Hon Hai, Quanta) & those taking hyperscale share (Quanta, Inspur). Morgan Stanley does and seeks to do business with companies covered in Morgan Stanley Research. As a result, investors should be aware that the firm may have a conflict Lower capex plans likely a function of growth slowdown and focus on of interest that could affect the objectivity of Morgan profitability. Lower capex coincides with slowing cloud & Internet service Stanley Research. Investors should consider Morgan provider revenue growth, suggesting reduced investments are tied to business, Stanley Research as only a single factor in making their investment decision. not temporal factors like technology cycles or reduced purchasing lead times FFoorr aannaallyysstt cceerrttiiffiiccaattiioonn aanndd ootthheerr iimmppoorrttaanntt ddiisscclloossuurreess,, (which was blamed for volatility in 2014). We also believe several companies rreeffeerr ttoo tthhee DDiisscclloossuurree SSeeccttiioonn,, llooccaatteedd aatt tthhee eenndd ooff tthhiiss in our tracker (e.g., Amazon, Google) are more focused on profitability, which rreeppoorrtt.. += Analysts employed by non-U.S. affiliates are not registered with FINRA, may could continue to pressure capex midterm. Some companies, including not be associated persons of the member and may not be subject to NASD/NYSE restrictions on communications with a subject company, public appearances and Seagate & QLogic, have suggested weaker cloud demand in the June quarter, trading securities held by a research analyst account. yet Intel highlighted strength in the cloud segment on earnings last week. 1

Global Technology| July 22, 2015 WWhhoo IIss MMoosstt EExxppoosseedd ttoo CClloouudd IInnffrraassttrruuccttuurree?? EExxhhiibbiitt 11:: Global Stocks Exposed to Cloud Infrastructure & Consensus Estimates US IT Hardware CY2015 Growth CY2015 Multiples US Software CY2015 Growth CY2015 Multiples Analyst: Katy Huberty Ticker Rating Revenue EPS P/E EV/Sales Analyst: Keith Weiss Ticker Rating Revenue EPS P/E EV/Sales Seagate STX UW -2% -8% 11.1x 1.2x Citrix CTXS EW 3% 10% 19.6x 3.6x Western Digital WDC EW -3% -3% 10.3x 1.1x MSFT MSFT EW 5% 3% 18.0x 3.4x HP HPQ OW -5% -1% 8.3x 0.6x VMware VMW EW 10% 12% 21.0x 4.4x QLogic QLGC EW 2% -2% 11.2x 1.3x Analyst: Jasmine Lu US Semiconductors Hon Hai Precision 2317.TW EW 7% 12% 10.1x 0.2x Analyst: Joe Moore Asia Semiconductors Cavium CAVM OW 17% 21% 36.2x 7.9x Analyst: Charlie Chan Inphi Corp IPHI EW 54% 94% 22.5x 3.1x Aspeed 5274.TWO EW 21% 23% 24.2x 7.3x Micron MU EW 0% -12% 7.1x 1.5x Inotera 3474.TW UW -11% -38% 4.7x 1.4x Altera Corp ALTR EW -9% -17% 40.0x 8.0x Asia Hardware Broadcom BRCM EW 2% 2% 17.6x 3.5x Analyst: Grace Chen SanDisk SNDK EW -16% -50% 20.1x 2.1x Inspur 000977-SZ OW 79% 96% 62.6x 2.8x Intel INTC UW -1% -4% 13.3x 0.2x Quanta 2382-TW OW 10% 10% 12.4x 0.3x Texas Instruments TXN EW 1% 7% 18.7x 4.1x Wistron 3231.TW EW 13% 14% 10.7x 0.1x Marvel MRVL OW -16% -43% 19.4x 1.3x Asia Hardware NVIDIA NVDA UW -3% -24% 25.4x 1.6x Analyst: Sharon Shin Xilinx XLNX EW -1% -1% 18.2x 3.8x Tripod Technology 3044.TW EW 2% -4% 10.9x 0.5x US Semiconductors Korea Technology Analyst: Craig Hettenbach Analyst: Shawn Kim Avago AVGO OW 49% 65% 15.3x 5.2x SK Hynix 000660.KS UW 14% 26% 5.7x 1.4x US Communications Systems Samsung Electronics 005930.KS OW 0% 1% 8.7x 0.8x Analyst: James Faucette Arista Networks ANET OW 39% 38% 42.3x 7.5x A10 Networks ATEN EW 8% 11% NA 1.6x Infoblox BLOX EW 20% 15% NA 4.1x Cisco Systems CSCO OW 4% 6% 12.5x 2.3x F5 Networks FFIV EW 11% 18% 17.8x 3.8x Juniper Networks JNPR EW -1% 25% 15.5x 2.1x Note: Consensus estimates. Source: Thomson Reuters, Morgan Stanley Research Please note that all important disclosures including personal holdings disclosures and Morgan Stanley disclosures appear on the Morgan Stanley public website at www.morganstanley.com/researchdisclosures. EExxhhiibbiitt 22:: Stock Performance of Cloud Exposure Companies versus Capex Trend Stock Performance Capex Growth Rate 160% 140% 120% 100% 80% 60% 40% 20% 0% -20% -40% 2010 2011 2012 2013 2014 2015 Source: Thomson Reuters, Morgan Stanley Research. Excludes Samsung from the share price performance given limited server exposure relative to significant market cap Companies with exposure to cloud infrastructure demand identified in Exhibit 1 historically trade in tandem with capex trends (Exhibit 2). In years where capex growth decelerated meaningfully (2010-12, 2015 YTD), the market cap for the group of stocks with cloud infrastructure contracted whereas in years capex growth remained stable (2013-14) market cap for exposed stocks expanded meaningfully (by 26-30%). We expect this correlation to hold going forward and see the deceleration in capex growth in 2015 continuing to weigh on share price performance. 2

Global Technology| July 22, 2015 CChhaannggeess ttoo MMoorrggaann SSttaannlleeyy EEssttiimmaatteess EExxhhiibbiitt 33:: Cloud Weakness Drives CY2015 MS Estimate and Rating Changes Morgan Stanley Estimate Changes CY2015 Revenue Growth CY2015 EPS Price Target Stock Rating New Old Chng New Old Chng New Old Chng New Old US IT Hardware HP -6.4% -6.2% 0% pts $3.58 $3.66 -2% $41 $42 -2% OW OW QLogic -3.2% -0.7% -3% pts $0.90 $0.96 -6% $12 $14 -14% EW OW Seagate -11.6% -10.1% -1% pts $3.46 $3.74 -8% $34 $49 -31% UW OW Western Digital -11.2% -5.6% -6% pts $6.41 $7.79 -18% $70 $113 -38% EW OW US Communications Cisco* 3.7% 4.1% 0% pts $2.22 $2.23 0% $30 $30 0% OW OW US Semiconductors Broadcom 1.8% 2.4% -1% pts $2.91 $2.95 -2% $58 $58 0% EW EW Cavium 14.8% 16.4% -2% pts $1.69 $1.76 -4% $68 $68 0% OW OW Marvell 6.2% 7.6% -1% pts $0.87 $0.94 -8% $17.50 $17.50 0% OW OW Avago 36.0% 37.8% -2% pts $9.05 $9.13 -1% $160 $160 0% OW OW Asia Hardware Tripod -1.9% 0.9% -3% pts NT$4.60 NT$4.96 -7% NT$52.50 NT$56.50 -3% EW EW Asia Semiconductors Aspeed 19.7% 23.0% -3% pts NT$12.93 NT$13.45 -4% NT$340 NT$350 -3% EW OW Source: Morgan Stanley Research. Cisco* estimates adjusted last week 3

Global Technology| July 22, 2015 CCyycclliiccaall aanndd SSeeccuullaarr PPrreessssuurreess oonn SSeerrvveerr SSppeenndd Our June Morgan Stanley CIO survey (“OOvveerraallll IITT BBuuddggeett GGrroowwtthh SSuussttaaiinnss,, BBuutt aa FFeeww CCrraacckkss SSttaarrttiinngg ttoo SShhooww”) reported overall enterprise IT budget growth expectations remained largely unchanged from our April survey. However, some cracks have appeared with higher expectations for budget cuts, a growing mentality of cost cutting, increased discounting, and longer purchasing cycles. Given the increased potential for downward IT budget revisions (Exhibit 3), we see the follow-through of cuts to overall IT budget growth as an elevated risk in our next survey in October. While cloud growth remains positive, a combination of slower revenue growth at some of the large hyperscale vendors and increased efficiency initiatives is pressuring spend in that market as well (detailed below). Overall we see: 1) potential for lower enterprise IT budget growth, 2) pressure on traditional OEM storage and server demand, and 3) slower cloud spending growth rate as material headwinds for data center infrastructure providers and their suppliers. EExxhhiibbiitt 44:: Revision Expectations Highlight Downside Risk to Enterprise IT Budget Outlook Source: Morgan Stanley CIO Survey 4

Global Technology| July 22, 2015 IInnttrroodduucciinngg MMoorrggaann SSttaannlleeyy CClloouudd CCaappeexx TTrraacckkeerr MMeetthhooddoollooggyy We have identified 19 companies that we believe have the highest predilection for aggressive IT infrastructure investments. We started with the highest ranked Internet companies by traffic, and included other large IaaS and SaaS providers. We also included Apple as it operates iTunes and cloud-based services to its user base as well as large US telco services providers (AT&T and Verizon). Our analysis incorporates total company capex including investments beyond building out cloud infrastructure and big data analytics. We included capex related to capital leases and backed out real estate investments where material. Our estimates are based off a combination of company guidance and Morgan Stanley Research estimates. We believe the companies in our tracker account for more than 40% of web traffic and cloud computing share. WWeebbssiittee TTrraaffffiicc DDrriivveess CClloouudd CCaappeexx TTrraacckkeerr Google (MS rated EW), Facebook (OW), YouTube, Baidu (OW), Yahoo, (OW), and Amazon (OW) are the top 6 websites ranked by global traffic in 2015 with Amazon ranking increasing most over the last year. EExxhhiibbiitt 55:: Top 20 Global Internet Domains by Traffic Position Traffic Year Ago Change from Rank Rank Year Ago Website Description Company 1 1 = google.com Search Engine, OS and Platforms Google 2 2 = facebook.com Social Networking Service Facebook 3 3 = youtube.com Video Sharing Service Google 4 5 +1 baidu.com Chinese Search Engine Baidu 5 4 -1 yahoo.com Search Engine Yahoo! 6 12 +6 amazon.com Online Marketplace, IaaS Provider Amazon 7 6 -1 wikipedia.org Web Encyclopedia Wikipedia 8 7 -1 qq.com Chinese Internet Service Portal Tencent 9 11 +2 twitter.com Microblogging Service Twitter 10 8 -2 taobao.com Chinese Online Marketplace Alibaba 11 14 +3 google.co.in Chinese Version of Search Engine Google 12 10 -2 live.com Web Application Service Microsoft 13 13 = sina.com.cn Chinese Infotainment Web Portal Sina 14 9 -5 linkedin.com Professional Networking Service LinkedIn 15 17 +2 weibo.com Chinese Microblogging Service Sina 16 19 +3 yahoo.co.jp Japanese Version of Search Engine Yahoo! 17 NA NA google.co.jp Japanese Version of Search Engine Google 18 24 +6 ebay.com Online Marketplace eBay 19 20 +1 yandex.ru Russian Search Engine Yandex 20 15 -5 blogspot.com Web Blogging Service Google Source: Alexa Internet, an Amazon company (as of July 2014). Note: The rank is calculated using a combination of average daily visitors to the site and page views on the site over the past 3 months. While not explicitly stated, we believe mobile traffic is not captured in Alexa rankings. 5

Global Technology| July 22, 2015 EExxhhiibbiitt 66:: Companies Included in Our Capex Tracker Y/Y Growth (%, on local currency) 2012 2013 2014 2015e AMZN 79% 16% 68% 12% AAPL (ex retail capex) 137% -31% 62% 18% BABA NA 40% 214% 39% BIDU 31% 19% 76% -11% CRM 16% 70% -3% 21% EBAY 30% -1% 2% 6% FB 46% -13% 33% 48% GOOG -5% 117%* 42%* 16% LNKD 41% 122% 36%* 39% MSFT 30% 96% -7% 29% QIHU 322% 72% 44% 19% RAX -2% 38% -6% 1% SINA -3% 83% 4% 2% T -3% 8% 1% -16% TENCENT -2% 23% -15% 54% TWTR 225% 44% 48% 48% VZ 0% 3% 4% 4% YHOO -15% -33% 10% 28% YNDX -28% 23% 94% 18% Average Capex Growth Rate 50% 37% 37% 20% Source: Company Data, Morgan Stanley Research estimates. SINA 2015 capex estimate per ThomsonReuters. Includes capital leases where material. *Reflects real estate capex backed out. 6

Global Technology| July 22, 2015 RRoobbuusstt CCaappeexx GGrroowwtthh SSlloowwss 1177 ppooiinnttss iinn 22001155 ttoo 2200%% Following an average capex growth rate of 37% in 2014, we expect the pace of capex spend to decline to 20% in 2015 (Exhibit 7). Beginning in 2012, the capex growth rate showed stabilization following aggressive investing for growth by Amazon, Apple, Baidu, Facebook, Tencent, and Yandex in 2010-11, ranging from 37% to 50% over the last three years. While results from our recent MS CIO survey keep us confident that cloud investments remain a priority for CIOs, we believe hyperscale (cloud + internet + large telco) data center spend is likely to decelerate as result of slowing revenue growth (Exhibit 8) and an increased focus on profitability. What's more, 2015 capex growth was revised lower, from 24% in 1Q15 to the current 20% forecast. EExxhhiibbiitt 77:: Average Capex Growth Decelerating to 20% in 2015 140% 140% 120% 94% 100% 80% 50% 60% 37% 37% Estimate in 40% 1Q15 24% 20% Estimate 20% in 2Q15 0% 2010 2011 2012 2013 2014 2015e Source: Company Data, Morgan Stanley Research estimates EExxhhiibbiitt 88:: Decelerating Capex Spend Likely Tied to Slower Revenue Growth Average Capex Growth Rate Average Revenue Growth Rate (RHS) 140% 70% 140% 63% 120% 60% 94% 100% 50% 43% 42% 35% 80% 40% 33% 50% 60% 30% 37% 37% 20% 40% 20% 20% 10% 20% 0% 0% 2010 2011 2012 2013 2014 2015e Source: Company Data, Morgan Stanley Research 7

Global Technology| July 22, 2015 Quarterly capex spend relative to the full year suggests 21% of investments occurred in 1Q15, compared to an average 19% in 1Q over the last three years (Exhibit 9). With 1Q15 off to a strong start, we believe risks for a large uptick in the back half of the year may be reduced, however current forecasts suggest 1H15 spend is in line with the three year average. Overall, we are incrementally cautious heading into the historically strong 3Q for cloud-related investments. We believe cloud spending will provide less of a cushion as it has in past years to offset weak spend in other areas including enterprise and consumer client segments in the near term. EExxhhiibbiitt 99:: 1Q Spend off to Strong Start Qtrly Capex as % of Full Year Capex 50% 43% 43% 40% 30% 21% 19% 20% 10% 0% 1Q 1H 2015 Estimate Trailing 3-Yr Average Source: Company Data, Morgan Stanley Research We capture ~$85B in capex spend at 19 cloud or consumer Internet service providers, which incorporates total company capex including investments beyond building out cloud infrastructure and big data analytics. While the split between companies expecting to accelerate/decelerate spending is roughly even (10/9 out of 19), the magnitude of the increase in 2015 is less than half the level in 2014. We believe slowing revenue growth and increased focus on profitability are likely to result in decelerating data center spending growth near- term. Below are examples of recent company commentary: Google (covered by Brian Nowak, EW) highlighted sequentially lower capex reflecting a digestion period in data center spend on the company's June quarter earnings call. Additionally, management also pointed out increased efficiency of new systems, with the ability to deliver three times as much compute power per capex dollar today versus five years ago on the March quarter earning conference call. Microsoft 's (covered by Keith Weiss, EW) June quarter capex grew sequentially in support of the company's cloud business. However, OpEx came in largely flattish Y/Y in FY15 despite continuing to aggressively invest, suggesting the company's focus on growing the business without incurring incremental costs. Amazon (OW) intends to deploy sufficient amount of capital to support accelerating growth at AWS, but has not provided specific guidance. We note that in May 2015, Amazon announced it will invest $1.1 billion in cloud data centers and one or more fulfillment centers in Ohio. However, the company highlighted a focus on productivity and efficiency efforts on the earnings conference call in January, which may result in a more managed approach to spending in the near term. Apple (OW) reiterated on the March quarter earnings conference call a $2B plan to build two data centers in Europe. However, we note guidance for $12.4B capex in 2015 (excluding retail 8

Global Technology| July 22, 2015 capex) reflects less than $2B incremental spend and 18% Y/Y growth, compared to $4B incremental capex during 2014 and 62% growth. Decelerating capex spend Y/Y has begun to weigh on server and storage spend, as noted by company commentary below. Western Digital (EW from OW) highlighted that cloud spending would undergo “periods of deployment, then rationalization, then deployment again” at the NASDAQ investor conference on June 30th. We believe this reflects a less optimistic tone compared to the March quarter earnings call (April 28th) in which the company signaled good visibility for cloud and “continued strong demand environment in the broader enterprise market, including hyperscale”. Seagate (UW from OW) previously expressed a similar less optimistic view on the April 17th earnings call by suggesting the high capacity drive market “can exhibit variations quarter to quarter as cloud service providers drive higher utilization of storage capacity additions and others prepare for buildouts in the second half of the year.” While the negative pre- annoucement this week was largely attributable to softer client and mission-critical enterprise, the company did signal less upside from cloud as in previous quarters. QLogic (EW from OW) indicated slower adoption of next-generation Grantley servers on its April 30th earnings call. We also attribute QLogic's negative June quarter preannouncement to broad weakness in the enterprise server and storage market while cloud demand is not providing the cushion it did in 2014. Aspeed (EW from OW) partially attributed lower June sales to data center customers hesitating to upgrade to the Intel Grantley platform, given the insufficient DDR4 server DRAM supply. Methode Electronics (NC) cited PowerRail fiscal 2015 sales (May 2 year-end) were approximately twice what the company anticipated as big data customer accelerated the build-out of its data centers. While this had a very positive effect on fiscal 2015, the company expects it will have a negative impact on fiscal 2016. However, we note Intel (OW) highlighted weakness in enterprise was offset largely by strength in the cloud on the company's earnings conference call last week. EExxhhiibbiitt 1100:: Sizable Cloud Exposure at QLGC, STX & WDC Source: Company Data, Morgan Stanley 9

Global Technology| July 22, 2015 UUSS IITT HHaarrddwwaarree:: DDoowwnnggrraaddiinngg IInndduussttrryy VViieeww ttoo CCaauuttiioouuss oonn SSlloowwddoowwnn iinn GGrroowwtthh Industry View: Cautious With the combination of weak enterprise and cloud infrastructure spend we lower estimates on HP, QLogic, Seagate, and Western Digital to better reflect the potential for below normal 2H15 seasonality. Without the potential upside from cloud, which was a key part of our thesis for STX, WDC, and QLGC, we downgrade stock ratings and our industry view from In-Line to Cautious. EExxhhiibbiitt 1111:: STX, WDC and QLGC Below Consensus Estimates by 8%, 5% and 6% CY2015 Consensus Revenue Over LTM vs New MS Estimate $18B $600M $17B $550M $15B $500M $14B $450M $12B $400M 4 4 4 4 4 4 4 4 4 4 4 4 4 5 5 5 5 5 5 5 5 5 5 5 5 5 5 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 / / / / / / / / / / / / / / / / / / / / / / / / / / / 4 8 1 5 9 2 6 0 4 7 1 5 9 2 6 0 3 7 3 7 0 4 8 2 5 9 3 / / / / / / / / 1 1 2 1 2 1 2 2 1 1 3 1 2 1 2 1 2 2 1 7 / 8 / / / / / / 1 / 2 / 1 / / / / / / / / 5 / 6 / 7 7 8 8 9 9 0 0 1 1 1 2 1 1 2 2 3 3 4 4 5 6 1 1 1 1 STX WDC QLGC (RHS) Source: Thomson Reuters, Morgan Stanley Research HP (OW, $41 PT): We lower server and storage revenue estimates to normal or below normal seasonality in each of the next four quarters. This reduces FY15e EPS by $0.05 to $3.55 and FY16e EPS by $0.11 to $3.68. While weak enterprise revenue near-term limits valuation upside, the pending separation remains a catalyst for unlocking value which keeps us OW the name. On the back of modest estimate cuts, our PT falls to $41 from $42, reflecting 11x FY15e EPS of $3.68 and in-line with other growth-challenged enterprise hardware stocks in our coverage. QLogic (OW to EW, $12 PT): QLogic is a direct play on the enterprise server and storage market and cloud server market via Fibre Channel Adapters which account for 67% of revenue and ship into enterprise and Ethernet products which account for 25% of revenue this year and ship into both enterprise and cloud servers. On the back of enterprise budget pressures, traditional server share losses to public cloud, and slower cloud capex growth, we lower FC Adapter and Ethernet revenue forecasts to reflect more normal to below normal seasonality. Our prior estimates called for a recovery off the June quarter weakness. This takes CY15 revenue growth to -3% Y/Y, from -1% previously, and EPS to $0.90, from $0.96 previously. Our $12 PT reflects both our lower EPS and a lower mulitple - 13x or 3-year average vs. our previous assumption that Win2003 and Grantley related server upgrades would push P/E to the high-end of the historical range, or 15x. With our new PT pointing to limited upside to shares, we downgrade the stock from Overweight to Equal-weight. Seagate (OW to UW, $34 PT): On the back of slower cloud capex growth, we lower our CY16e high capacity cloud unit growth forecast to 8% for the HDD industry down from an average of 18% over the last three years. 10

Global Technology| July 22, 2015 Seagate also generates a large portion of its demand from cloud customers in China which we view as an incremental risk to estimates given recent stock market volatility. The combination of less favorable mix and a continued drag from the Xyratex business will pressure margins to the low-end of the company's long-term range in FY16, with potential for more downside if enterprise/cloud remains weak for an extended period of time. As a result of our lower estimates, we move to an Underweight rating (from Overweight) with a $34 price target which applies the 3-year average multiple or 10x to our FY16 EPS estimate of $3.38. WD (OW to EW, $70 PT): Incorporating new cloud HDD forecast highlighted above and incremental enterprise weakness highlighted by our CIO survey lowers our FY16 EPS estimate to $6.35 from $8.66. Volume declines and less favorable mix are only partially offset by strong growth in WDC's emerging flash business causing us to reduce our gross margin forecast to 28.7% from 31%. Weakening demand and cost pressures may increase the likelihood that WDC receives approval from MOFCOM to fully integrate HGST, adding as much as $1.50 EPS upside. Net, we move to EW with a $70 price target from OW (prior price target was $113) given the less favorable mix and EPS pressure. However we apply a slightly higher EPS multiple to WDC (11x) vs. STX (10x) given the combination of WDC's faster growing flash business and potential MOFCOM approval. EExxhhiibbiitt 1122:: Changes to Morgan Stanley Estimates FY2016 Revenues ($M) FY2016 EPS New Old Change vs Cons New Old Change vs Cons HP $99,741 $101,657 -2% -4% $3.68 $3.79 -3% -3% QLogic $476 $495 -4% -7% $0.84 $0.94 -10% -16% Seagate $12,354 $12,805 -4% -6% $3.38 $4.08 -17% -20% Western Digital $13,296 $14,635 -9% -6% $6.35 $8.66 -27% -15% Price Target Stock Rating New Old Change vs Close New Old HP $41 $42 -2% 35% OW OW QLogic $12 $14 -14% 4% EW OW Seagate $34 $49 -31% -29% UW OW Western Digital $70 $113 -38% -11% EW OW Source: Thomson Reuters, Morgan Stanley Research 11

Global Technology| July 22, 2015 UUSS SSeemmiiccoonndduuccttoorrss Industry View: In-Line Summary: Cloud weakness should represent a headwind in 2h to infrastructure names already facing weakness in wireless infrastructure and weaker enterprise spending. We reiterate our below consensus views on Intel (where we expect data center growth of 10-12% vs. company forecast 15%) and server memory, and trim our estimates for Avago's cloud related business (offset by wireless increases), Broadcom, Cavium, and Marvell; we note that the absolute earnings impact is relatively small, but that this adds to overall headwinds and higher risk of disappointment. Details: Within semiconductors, data center infrastructure has been a growth segment in a sector struggling with growth concerns in verticals such as PCs, smartphones/tablets (ex Apple), and wireless infrastructure. This has happened despite headwinds in service provider spending on wireline, and more recently enterprise growth concerns. EExxhhiibbiitt 1133 shows our estimate of total semiconductor exposure to cloud, which overall remains relatively small. Still, deceleration in this segment could prove problematic to 2nd half numbers, as there simply aren't enough other growth drivers to compensate, in some cases. Intel - Cloud about 6% of revenues - maintaining below consensus server #s. Total Data Center Group (DCG) is about 29% of revenues, Intel has indicated in several public reports that cloud represents about 20% of DCG, with the balance being enteprise IT (45% of revenues), telco/service provider (10% of revenues), and government/academia/science. We have already budgeted for further deceleration in DCG, which is part of the reason that we are below consensus. We like Intel's virtual monopoly position in DCG, and see average selling prices rising over time, but have generally thought that Intel is a bit too optimistic in calling for a 15% organic growth rate. We note that the company has only achieved this growth rate once in the last three years, in 2014, in a year that was enhanced by very strong cloud buildout. Growth was 6% in 2012, and 7% in 2013, in a strong economic environment with a heavy cloud buildout and penetration of telecom/service provider. The company also acknowledged seeing some strength from replacing Win2003 servers last year, as Microsoft's support expires in July of this year. In particular, we had highlighted after 4Q14 that the 25% y/y growth rate that the company reported was likely not sustainable. While Intel has been qualitatively positive on DCG YTD, we note that 1Q14 was the worst sequential comparison for the segment in four years (-10%), and that Y/Y growth has already decelerated to 10% in 2q. We estimate only 5% q/q growth in 3q, and 4% in 4q, bringing full year growth to 10% vs. company guidance of >15%; further, we estimate 2016 revenues at 12%. We note that long term, the growth in cloud is a net negative for Intel's DCG business, as cloud vendors will manage much higher utilization rates than enterprise data centers. Broadcom: Cloud exposure 7% of revenues, trimming estimates slightly to 4% Infrastructure and networking growth this year. The Infrastructure and Networking segment is 31% of Broadcom; within that, the company has described roughly 3 equal subsegments - carrier networks, data center networks, and campus networks - and within data center networks, we estimate that roughly 2/3 of the business is public cloud. We note that our numbers already take into account substantial deceleration in the business, due to the slower carrier spending that we have seen in the last three quarters. Before this cut, we projected infrastructure and networking growth this year of 7%, after 17% growth last year. Assuming a slower cloud spending year in line with this report, we take our 3Q sequential revenue growth for I&N from 5% to 1%, and assume a flat quarter in 4q. This takes total full year N&I revenue growth down to 4%. We continue to look for 11% growth next year; we see a significant tailwind to the BRCM #s from better long 12

Global Technology| July 22, 2015 term penetration of Cisco's switching business, which has historically been mostly served by custom silicon. Refer to BBrrooaaddccoomm:: AA cclloosseerr llooookk aatt CCiissccoo ooppppoorrttuunniittiieess ffoorr BBrrooaaddccoomm ((0088 OOcctt 22001144)) . Cavium: Cloud exposure 15% of revenues, though very narrow (Amazon); we trim numbers slightly. We model Cavium's total enterprise infrastructure exposure (vs. service provider) at just over 50% of revenues this quarter. But of this, most of the revenue comes through enterprise-centric customers such as Cisco (just below 20% of revenues), F5, Citrix, and Palo Alto Networks. Cavium does have ~10% exposure to Amazon, through its Liquid I/O business. We are budgeting for a sharp deceleration in Amazon's capex this year, but do not think that the Liquid I/O business can be triangulated through that. So Cavium's absolute exposure to cloud is relatively small and narrow. Having said that, most of the company's key growth initiatives - Thunder, xPliant, and Liquid I/O 2, are largely dependent upon early adopters within cloud, so we do think there are headwinds. Overall we are trimming our enterprise expectations by a couple of percent for both 3q and 4q. Marvell: Cloud exposure 7% of revenues, across both storage and networking; we trim numbers slightly for the October and January quarters. Marvell's enteprise exposure comes both from the enterprise portion of its storage business (mostly HDD, but also some SSD including custom controllers for SNDK Smart), and from its networking business. We still expect a bounceback in storage in F3q and F4q, given the overcorrection in 1h16, but have taken a few percent out of our October quarter forecast for that segment. Inphi: Cloud exposure about 10% of revenues, mostly memory and Cortina. Inphi has already been faced with headwinds in this business, given the share loss in the DDR4 segment of memory buffers; we believe that total cloud exposure is 20% or so of memory, 20% of the Cortina 10/40G Phy business, and a small portion of the core communications business. Similar to Cavium, we believe that a large portion of the growth story for 2016 and beyond would come from cloud, so long term trends are still important. We are not changing estimates here, as we had already anticipated a 3% decline in server memory in 2q, and a 12% decline in 3q, given already understood share challenges. FPGAs (Altera and Xilinx): Cloud exposure about 2-3% of revenues. We think FPGA data center exposure is mostly server acceleration, and enterprise SSD. We do think that the data center businesses skew a bit more towards hyperscale than other silicon suppliers, given the faster pace of hyperscale This adds to a fairly long laundry list of concerns for FPGAs in the current environment, though the absolute impact is at the noise level. Memory (Micron and Sandisk): Cloud exposure: MU 5% of revenues, SNDK about 2%; no change to our cautious outlook on server DRAM and NAND. We believe that cloud reflects about one-third of Micron's server DRAM, which is about 18% of total bits, and about 4% of total NAND, driving about 5% of revenues overall. We are below consensus on server DRAM pricing, as we think that both DDR3 and DDR4 have continued to come under incremental pressure. The primary issue is the oversupply in DRAM overall impacting pricing, as opposed to weaker server demand, but a shortfall in demand wll not help. Recent checks show both DDR3 and DDR4 pricing well below the prices published on DRAMexchange, which themselves show significant declines. For NAND supply and demand, we are somewhat more optimistic. But we note that over time we would expect the systems margins on enterprise SSD to come down, and think this is key to the NAND vendors taking share. We note that Sandisk and Micron's exposure to enterprise SSD remains relatively small. Avago: Cloud exposure is about 7% of revenues. Avago’s exposure to the cloud is mostly centered in its Enterprise Storage segment through its HDD SoC and Server & Storage connectivity business and Emulex. In addition, the company has some exposure to the cloud in its Wired Infrastructure segment through the sale of switching ASICs and fiber optics to hyperscale vendors. Given we see some headwinds for Storage and Wired Infrastructure segments from a slowdown in cloud spending, we trim our July and October quarter growth estimates in Storage to 23%/2% from up 25%/4% previously and in Wired Infrastructure to -5%/+4% from - 3%/6%. We had thought our wireless estimates were likely conservative (up 8% q/q in the July quarter and 15% in the 13

Global Technology| July 22, 2015 October quarter), although with iPhone unit shipments coming in slightly below expectations that may limit the potential upside. That said, we are still very positive on Avago's opportunity to increase $ content in wireless, which should be further aided by an expansion in FBAR capacity. Finally, despite some potential headwinds related to the cloud, we expect the company to manage opex aggressively to protect EPS and point out opportunities to reduce costs at Emulex. Overall our C15/16 Sales estimates come down by 1%/2% to $7.07bn/$7.73bn and EPS estimates come down by $0.08/$0.16 to $8.26/$9.25. The stock is trading at only 13.7X on our CY16 EPS (including stock based comp) and 11X on run rate EPS with BRCM. EExxhhiibbiitt 1133:: US Semiconductor Server and Cloud Exposure as % of Revenue Source: Morgan Stanley Research estimates 14

Global Technology| July 22, 2015 UUSS CCoommmmuunniiccaattiioonnss SSyysstteemmss Industry View: Cautious For most of the networking companies that we have under coverage, our 2H15 estimates are below consensus. Recent news flow and checks have increased our confidence in those below-consensus estimates on Arista, A10 Networks, F5 Networks, and Infoblox. Our conservative outlook stems from two key sources: 1. Expectations that product cycle-driven strength would begin to abate during 2H15, particularly ahead of new data center switching products due late in 2015 and early 2016. 2. Our view that the hyperscale web properties were spending to add capacity well ahead of their need to deliver it. While we haven’t known for sure when that scale build out phase would begin to abate, we believed it prudent to start to build a related slowdown into this year’s estimates. Perhaps we are beginning to see our suspicions confirmed. On the other hand any slowdown may be more a function of an aforementioned pause in spending ahead of new switching products combined with the front-half 2015 weighted IT spending and increased security project focus highlighted in our most recent CIO survey rather than any new spending philosophies. For A10 Networks, F5 Networks, and Infoblox, we remain comfortable with our estimates. We also like our 2015 Arista estimates, but as outlined in our note “PPrroodduucctt RReeffrreesshh PPootteennttiiaall LLaatteerr TThhiiss YYeeaarr aass TToommaahhaawwkk--BBaasseedd PPrroodduuccttss RReelleeaasseedd” from March 26, 2015, we believe that successful execution on new switch launches based on the Broadcom Tomahawk chip, and potentially other suppliers, could drive meaningful upside to our 2016 estimates on a stronger-than-expected upgrade cycle. For Cisco, we believe that demand is quite healthy, and the company continues to benefit from strong response to its newest switching and routing products. Nevertheless, in our QQ22 NNeettwwoorrkkiinngg PPrreevviieeww we reduced estimates slightly to take a more conservative view on October quarter estimates, particularly as that will be the 1st quarter of the company’s new fiscal year and we believe we had previously modeled seasonality incorrectly. For Juniper, we are slightly above consensus for the coming few quarters as we believe the company is seeing good response to its new QFX core switch and vMX virtual routing products. There may be a bit of risk to our estimates from Juniper’s data center and enterprise segments (about 1/3rd of revenue), but in the near term we believe the stock and our estimates reflect a positive reception for the company’s newest products. 15

Global Technology| July 22, 2015 EExxhhiibbiitt 1144:: CSCO: Changes to our estimates - July 17, 2015 Source: Morgan Stanley Research 16

Global Technology| July 22, 2015 UUSS SSooffttwwaarree Industry View: In-Line Citrix (CTXS.O, EW): Citrix's primary exposure to cloud service providers comes via its Netscaler business which we estimate was 18-20% of total revenue in 2014. Approximately one-third of Netscaler sales is to a handful of large cloud providers who have infrequent but large order patterns. The lack of large orders in recent quarters has negatively impacted Netscaler performance in recent quarters and, given the expectation for lower cloud capex growth, the benefit Citrix realizes when these orders return may be less than expected. Therefore, we see some risk to our NetScaler estimates, but minimal risk to our EPS estimates as the company is in cost- cutting mode as it implements the restructuring plan announced in December 2014. Furthermore, the potential for value creation from activist shareholder involvement, which involves further actions to reduce costs, may be more likely to determine Citrix's share price performance in the near term. 17

Global Technology| July 22, 2015 AAssiiaa HHaarrddwwaarree Industry View: In-Line We are more constructive on Asia Hardware in light of significant exposure (~45-50%) to Apple where we see iPhone continuing to drive upside versus consensus estimates. On the other hand, server exposure is lower (~10-15%) which is where we see incremental risk if cloud data center capex slows, and their share gains should help offset the slower demand. Below we highlight companies with exposure to server and storage demand. Generally speaking, we're more positive on these names than our US counterparts given growing market share relative to US server OEMs. Quanta (2382.TW, OW): We maintain our OW rating on Quanta in view of its expansion in the datacenter hardware business, Apple Watch potential, and undemanding valuation (11x 1yr-forward EPS) with a 6% cash yield. Quanta’s datacenter hardware sales is on track with our expectation to grow 20-30% YoY in 2015, despite the market’s concerns about server demand slowdown. We attribute Quanta’s outperformance to 1) expanding product offerings to rack-level solutions from server boards/systems only; and 2) the expansion into the enterprise segment. Nonetheless, we expect its OPM for datacenter hardware may fall YoY in 2015 due to a higher sales mix of rack solutions as they yield a lower margin rate but higher profit dollar due to much higher ASPs than standalone servers. While there is debate about Apple Watch demand, we expect the margin for this project will be gradually improving for Quanta after it started mass production in the June quarter. Inspur (000977.SZ, OW): We have an OW rating on Inspur. Inspur’s server business remained on track with our forecast despite the market’s recent concern about a demand slowdown. Inspur’s server sales rose ~60% YoY in 1H15, which is on track to reach our full-year forecast of 63% YoY. Datacenter servers remained the key driver. In addition, we believe Inspur may likely benefit from share gain at the expense of other server brands. Hon Hai Precision (2317.TW, EW): Hon Hai generates roughly 10% of sales from server and storage products. We believe low-end, cloud-related shipments via partnership with HP could be around US$1 billion. These include sales into Microsoft and Chinese OEMs. Tripod (3044.TW, EW): Tripod has 35% revenue exposure to PC/NB segment and another 14% revenue contribution from server PCB. The demand slowdown at both server and PC segment will likely drag Tripod’s business momentum in 2H15 and 2016. Thus, we revised down our 2015-17e earnings estimates for Tripod by 7-9% and thus, lower our price target to NT$52.5, implying 11.4x 2015e PE. We retain our Equal-weight rating as we see its growing exposure at automotive segment should help mitigate some business shortfall at PC/server segment. Tripod trades at 12x 2015 PE, fair to its historical range of 8-15x or peers’ 9x-13x. 18

Global Technology| July 22, 2015 AAssiiaa SSeemmiiccoonndduuccttoorrss Industry View: In-Line Aspeed (5274.TW, OW to EW): The company generates 90% of its revenue from sever BMC (board management controller). It has 25% global market share in the BMC chip, key end-customers are data center vendors such as Microsoft, Facebook, Amazon, and China B.A.T. The company has won several Grantley projects this year, such as from Supermicro. However, Aspeed’s 1H15 revenue has been tracking below our estimates given a delayed server replacement cycle. There wasn’t any market share or ASP issue, so the revenue shortfall was purely demand-driven. With more evidence that the server demand will not recover sharply in 2H15, we cut our 2015e and 2016e EPS forecasts by 4%-5%. The stock now doesn’t look attractive at 26x P/E vs. 21% EPS growth in 2015e and we lower our rating to EW, from OW. Inotera (3474.TW, UW): Server DRAM accounts for 50% of Inotera’s sales in 2015e. The demand shortfall of server DRAM would enlarge the gap of DRAM oversupply in 2H15 and 2016e. We have been factoring very conservative DRAM price assumption for Inotera based on our bearish view on the DRAM cycle. We therefore maintain our earnings forecasts but reiterate UW rating on the stock, given its heavy exposure to sever DRAM. 19

Global Technology| July 22, 2015 KKoorreeaa TTeecchhnnoollooggyy Industry View: In-Line SK Hynix (000660.KS, UW): The company generates around 20% of revenues and ~25% profits from server DRAM. The 20% premium for DDR4 products over DDR3 at the beginning of the year has eroded to 10% in Q2 and we expect to quickly disappear entering Q3. We have recently lowered our DRAM revenue estimates to reflect a shortfall in demand growth from the server market in each of the next four quarters. We continue to have significant fundamental concerns, as we think PC DRAM oversupply has spread to both servers and mobile. SK Hynix is starting to price in the reality of weaker end-demand, but DRAM prices remain under pressure in Q3, which means underlying downgrades are likely to continue and sustain the recent share weakness. We are well below consensus, which keeps us UW the name. For Samsung Electronics (005930.KS, OW), the impact is less meaningful as exposure to server is ~3% of revenues with the bulk for its own mobile IM division and the stock is largely driven by mobile market share and margins. 20

Global Technology| July 22, 2015 LLoowweerriinngg HHPP EEssttiimmaatteess Lowering Estimates for Hewlett-Packard (HPQ), Remain OW with $41 PT, down from $42 While HP has heavily invested in the server and storage product portfolio over the last several years, including refreshing the ProLiant and Apollo server lines, introducing HP Moonshot, and driving continued adoption of converged solutions and 3PAR, we lower server and storage revenue estimates to normal or below normal seasonality in each of the next four quarters to reflect broader market weakness in both the traditional enterprise and cloud. As a result, our FY15 EPS decreases by $0.05 to $3.55 and FY16 EPS by $0.11 to $3.68. While weak enterprise revenue, potential for further one-time separation cost announcements, and overhang from M&A speculation (e.g., Bloomberg June 3, 2015) could limit valuation upside near-term, the pending separation remains a catalyst for unlocking value, which keeps us Overweight the name. On the back of modest estimate cuts, our PT falls to $41, from $42, reflecting 11x FY16 EPS of $3.68, in-line with other growth- challenged enterprise hardware stocks in our coverage. 21

Global Technology| July 22, 2015 HHPP RRiisskk RReewwaarrdd IInnvveessttmmeenntt TThheessiiss $ 60 We see headwinds weighing on FY15e FCF largely reversing in FY16e, driving a more normalized 50 $50.00 (+64%) performance. Low investor expectations headed into the split and potential for multiple expansion $41.00 (+35%) 40 set up for attractive return post-split, in our view. $30.45 30 $30.00 (-1%) KKeeyy DDeebbaatteess 20 How does the announced split unlock value? We see a combination of low investor expectations and 10 multiple re-rating setting up for attractive return post-split as 1) HP Inc. turns into a yield vehicle and therefore re-rates toward peers with 0 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 significant yield, and 2) services margin expansion Price Target (Jul-16) Historical Stock Performance Current Stock Price WARNINGDONOTEDIT_RRS4RL~HPQ.N~ in Enterprise provides further evidence of a Source: Thomson Reuters, Morgan Stanley Research successful turnaround. Our updated SOTP points to $47-53, 40-60% upside from current levels. Derived from the base-case scenario. Price Target $41 What is HP’s revenue growth trajectory? We model FY15 revenue to come in roughly flat versus Clearly outlined separation path drives shares toward SOTP Bull $50 FY14 in constant currency, with limited revenue valuation . Our analysis suggests shares could converge toward SOTP Valuation disruption caused by the announcement to split $47-53 implied in SOTP valuation. The company executes on new HP. cost reduction opportunities to offset any dissynergies in the Could HP return to 10-11% operating margin? separated businesses. HP Inc. is perceived as a yield vehicle post- Longer-term, yes. We model 10bps operating split and therefore re-rates toward peers with significant yield, margin expansion in FY15 despite currency while strength in servers and services margin expansion in headwinds and expect margin to expand to 9% in Hewlett-Packard Enterprise provides further evidence of successful FY16, with further improvements likely in the outer turnaround. years. Currency headwinds and separation costs weigh on EPS and Base $41 PPootteennttiiaall CCaattaallyyssttss FCF in FY15, but performance mostly normalizes in FY16. 11x FY16e EPS of $3.68 / 13x Operating margin expands modestly despite increased standalone FCF $3.18 Lower than expected disruption from the company costs - to 8.9% in FY16, up from 8.8% in FY15 and FY14, announced split driven by Enterprise Services margin expansion. Free cash flow Better Enterprise Services margins following comes in at $3.18, up from $1.59 in FY15. 11x P/E is still well restructuring savings, lower account run-off, and below peer multiple range of 14-15x. renegotiated deals Early revenue and margin pay-off from Shares trade on FY15 earnings, with lower FCF conversion Bear $30 investments weighing on multiple. Steeper than expected currency headwinds 9x FY16e EPS/FCF $3.50 Greater balance sheet flexibility to invest in new and PC market demand deterioration results in FY15 EPS coming products and more aggressively repurchase shares in below the guided range, while weaker FCF conversion (largely a result of separation-related charges) weighs on P/E multiple causing shares to re-rate toward 9x. RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett Cost dissynergies drive discount to SOTP valuation PC spending weakness post the enterprise refresh cycle Continued need to restructure and invest in the product portfolio Competitive environment drives margin pressure 22

Global Technology| July 22, 2015 DDoowwnnggrraaddiinngg QQLLooggiicc Downgrading QLogic (QLGC) to EW, from OW, with $12 PT Downgrading to Equal-weight from Overweight: Our prior Overweight thesis was predicated on 1) Windows Server 2003 support expiration and Intel’s Grantley lifting enterprise server demand, 2) cloud demand providing upside to QLogic’s Ethernet business, and 3) P/E multiple sustainably re-rating toward the top of historical trading range on the back of cloud exposure, which QLogic gained with its Ethernet asset acquisition last year. Our CIO survey points to 1) weakening enterprise IT budget outlook and 2) more front-end loaded seasonality in 2015, both of which are negative for QLogic. Combined with the slowing cloud capex growth trends, we see limited upside to estimates and downgrade QLGC to Equal-weight from Overweight. Lowering PT to $12 from $14: We lower Fibre Channel Adapter and Ethernet revenue estimates to reflect more normal to below normal seasonality. Our prior estimates called for a recovery off the June quarter weakness. This takes CY15 revenue growth to -3% Y/Y, from -1% previously, and EPS to $0.90, from $0.96 previously. We lower our target P/E multiple from 15x (which was toward the high end of historical trading range) to 13x (the average of historical trading range) as weaker demand across enterprise and cloud takes sustainable multiple expansion potential off the table in the near-term and as a result, our price target decreases to $12 from $14 previously. Where we could be wrong – bull/bear case view: Our $18 bull case assumes enterprise demand shows a strong recovery in C2H15, driving operating leverage and upside to estimates. Bull case also assumes the Ethernet business shows robust recovery and can sustainably grow, pushing P/E multiple to 16x, which is at the top of historical trading range. Our bear case of $8 will materialize if estimates continue to see downside on the back of even weaker demand than we currently bake into our earnings forecast – either from reduced data center investment by enterprises (as they adopt cloud) or cloud data centers (as they become more efficient and see slowing revenue growth with scale). Our bear case assumes P/E drops toward the low-end of QLogic’s historical range, or 10x, similar to other periods of weak top-line trends. EExxhhiibbiitt 1155:: We Now Assume QLogic Trades In-Line with Historical Averages Source: Thomson Reuters, Morgan Stanley Research 23

Global Technology| July 22, 2015 QQLLooggiicc RRiisskk RReewwaarrdd IInnvveessttmmeenntt TThheessiiss $ 20 Near-term enterprise market weakness drives 18 $18.00 (+57%) revenue decline in CY15 while weaker hyperscale 16 data center spend limits upside optionality and 14 returns P/E closer to 3-year average of 13x. $11.48 12 $12.00 (+5%) 10 KKeeyy VVaalluuee DDrriivveerrss 8 $8.00 (-30%) What is QLogic’s revenue growth rate? We expect 6 revenue to decline 3% in CY15 on the back of near-term enterprise market softness. Longer- 4 term, we see risk to low single digit revenue 2 growth forecast if QLogic doesn't improve 0 exposure to the hyperscale market. Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 Price Target (Jul-16) Historical Stock Performance Current Stock Price WARNINGDONOTEDIT_RRS4RL~QLGC.O~ What is QLogic's long-term operating margin Source: Thomson Reuters, Morgan Stanley Research profile? While Ethernet business carries a lower margin today, disciplined actions to extract further Derived from the base-case scenario. Price Target $12 efficiencies in the cost structure, could improve margins but the long-term target of high-teen Enterprise demand rebounds in C2H15, driving strong non-GAAP operating margin may prove difficult Bull $18 operating leverage. Revenue grows double-digits in CY15 and without top-line growth. 16x P/E on CY15e non-GAAP non-GAAP operating margin comes in north of 20% on the back EPS of $1.15 of strong operating leverage. Our 16x P/E multiple assumes PPootteennttiiaall CCaattaallyyssttss QLogic shares re-rate towards the higher end of the three-year Weak enterprise and cloud demand in 2H15 trading range to account for increased investor confidence in long- term growth prospects. Product introductions expand QLogic’s TAM Potential cost cuts help offset revenue decline Near-term softness drives negative operating leverage while Base $12 weaker enterprise demand and limited upside potential from 13x P/E on CY15e non-GAAP RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett cloud lowers P/E versus recent range. Traditional Advanced EPS of $0.90 FC adapter market declines faster than expected as Connectivity Platform revenue declines 12% in CY15 versus 4% workloads continue to shift to the cloud decline in CY14 on the back of enterprise market weakness, while the Ethernet business acquired from Broadcom contributes over Price competition in 10 Gigabit Ethernet pressures $120 million in CY15. Non-GAAP gross margin comes in at 61.6% margins and operating margin contracts to 18%. Our 13x P/E multiple is HBAs replaced with chip-level solutions or other the average over the last three years and no longer reflects upside protocols from cloud-related Ethernet demand or a meaningful enterprise server cycle in 2H15. Downward estimate revisions continue. Windows Server 2003 Bear $8 support expiration, Intel Grantley and Ethernet drive less revenue 10x P/E on CY15e non-GAAP than expected. Revenue declines mid to high single digits in CY15 EPS of $0.80 and non-GAAP operating margin approaches mid-teens given lack of operating leverage and continued need for investments. As back half recovery hopes do not materialize, multiple is depressed toward the low-end of the historical trading range to reflect end demand challenges and investor concerns. 24

Global Technology| July 22, 2015 DDoowwnnggrraaddiinngg SSeeaaggaattee Downgrading Seagate (STX) to UW, from OW, with $34 PT Downgrading to Underweight from Overweight: Our prior Overweight view was that the combination of upside from cloud orders and price stability from a consolidated industry could drive gross margins higher over time. While the industry continues to behave rationally, with inventory at three-year lows and pricing stable, we now see less uplift from cloud. As a result of our lowered view on cloud capex (accounts for as much as 12% of revenue, higher percent of profits), we lower our FY16 (fiscal year-end June 2016) revenue and EPS to $12.4B and $3.38, from $12.8B and $4.08 previously. Our estimates are now well below consensus (by 6% for Revenue and 20% for EPS). We now believe Seagate will lose much of the cushion from large hyperscale/cloud storage purchases it enjoyed over the past year and was a key. We believe near-line or capacity optimized storage, which includes these cloud data center purchases, accounts for 14% of revenue (FY14) and a higher portion of profits. Additionally, slower unit growth in this category, combined with lower client and enterprise units, will pressure fixed cost absorption. Our lowered estimates still assume Seagate can deliver the low-end of its 27-32% gross margin target but weaker volume and enterprise/cloud mix could push it lower without significant restructuring over the next few quarters – making the bear case more likely than the bull case in the near-term. Lower PT to $34 from $49: Our prior $49 PT credited STX with multiple expansion from historical levels as cloud became a larger growth driver and inventory/pricing behavior reflected a more consolidated industry. We now believe that with weaker demand across PC, enterprise, and cloud that P/E is likely to return to the high-end of the three year average range of 8-10x. Our $34 PT is based on 10x our lowered FY16 EPS of $3.38 as compared to a 12x P/E implied by our prior target of $49. The new 10x P/E is now at the low-end of IT Hardware ranges, reflecting our view that top-line will decline by 10% in FY16 and EPS by 24% (more than revenue due to margin weakness from lower fixed cost absorption and fewer mix benefits). Where we could be wrong – bull/bear case view: Our bull case of $62 stock price assumes the weakness in cloud data center growth is short-lived and a recovery in unit growth pushes gross margin to the upper end of Seagate’s 27-32% target range. Cloud strength offsets PC/enterprise weakness and allows the company to deliver slight EPS expansion in FY16 despite meaningful investments in upgrading the product portfolio this year. Our bear case of $25 assumes gross margin breaks the 27-32% target range as weaker volumes and mix shift away from higher margin categories, enterprise and cloud, offset the benefits of a consolidated industry. 9x bear case EPS of $2.76 assumes P/E multiple returns to the mid-point of Seagate’s three year average range of 8-10x and in-line with the lower end of ranges for IT Hardware peers including HPQ. 25

Global Technology| July 22, 2015 SSTTXX RRiisskk RReewwaarrdd Seagate: Enterprise and Cloud weakness raises revenue growth and margin risk IInnvveessttmmeenntt TThheessiiss $ 80 Weaker enterprise IT budget and cloud demand 70 continue into 2H15, driving downward revisions to $62.00 (+29%) consensus estimates. P/E falls back toward high 60 end of the historical range of 8-10x pre- $48.11 50 consolidation. 40 KKeeyy DDeebbaatteess $34.00 (-29%) 30 Can Seagate achieve the upper-end of its $25.00 (-48%) 20 targeted gross margin range in CY15? Unlikely near-term given unit shortfalls, particularly in 10 enterprise and cloud markets. Longer-term, it will require faster adoption of adjacent cloud solutions. 0 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 While enterprise weakness presents risk of falling Price Target (Jul-16) Historical Stock Performance Current Stock Price WARNINGDONOTEDIT_RRS4RL~STX.O~ below the 27-32% target range, low channel Source: Thomson Reuters, Morgan Stanley Research inventory and stable pricing are potential offsets. What is the long-term unit growth outlook? We Derived from the base-case scenario. Price Target $34 model double-digit unit decline in FY16/CY15 due to the combination of weak PC, enterprise, and Unit growth and cloud mix increase margins to the 30% Bull $62 cloud data points. We currently forecast flattish range. June quarter cloud weakness proves short-lived and new 14x FY16e EPS of $4.46 units in FY17 but see risk if structural pressures high-capacity cloud and hybrid units provide margin tailwinds such as SSD and cloud adoption accelerate. while technology transitions prove less disruptive. The company exceeds $1B revenue run rate in non-HDD businesses in 2H15. PPootteennttiiaall NNeeggaattiivvee CCaattaallyyssttss Seagate trades just below the market multiple given the growth characteristics of the industry but toward the upper-end of its Cloud and Enterprise demand prove front-end two-year trading range. loaded, leaving estimates too high for the remainder of the year Incremental enterprise and cloud weakness suggest both Base $34 SSD adoption in client and enterprise systems eat estimate and multiple downside. Mission critical revenue decline 10x FY16e EPS of $3.38 into HDD market opportunity (-16% Y/Y) and slower cloud growth (+6% Y/Y) raise margin risk Mismatched supply/demand fundamentals and pressure P/E toward 3-year average of 9-10x. Our base case Higher market share targets at Toshiba assumes management manages manufacturing capacity to maintain gross margin at the low-end of the long-term target of Slowing areal density / technology transition 27-32% but we see potential risk to this assumption in our bear lowers margins case. Incremental investments to capture large cloud systems TAM Weaker demand and SSD threats cause gross margins to fall Bear $25 Tax rate increases below the low-end of the company’s long-term range. 9x FY16e EPS of $2.76 However, cash flow characteristics create a valuation floor. The RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett stock trades back within its long-term P/E range prior to industry consolidation given the unit pressure and reduced margin profile. Ramp of leading-edge products - high capacity cloud Stabilizing PC build rates Xyratex increases cloud growth OpEx savings from regulator approval for full integration of Samsung or structural cost improvements to better align with lower unit growth Increased share repurchases provide support to share price 26

Global Technology| July 22, 2015 27

Global Technology| July 22, 2015 DDoowwnnggrraaddiinngg WWeesstteerrnn DDiiggiittaall Downgrading Western Digital (WDC) to EW, from OW, with $70 PT Downgrading to Equal-weight from Overweight: Our prior Overweight view was that the combination of 1) upside from cloud orders, 2) benefits of a consolidated industry including more stable pricing, 3) better execution in flash storage with recent acquisitions, 4) potential upside from MOFCOM decision allowing for incremental cost reductions. While some of these catalysts are still in play – including industry consolidation and better execution outside of HDDs – the combination of lower demand from cloud customers and no evidence of a MOFCOM decision near-term make us believe earnings estimates and valuation multiples may be pressured near-term. As a result of our lowered view on cloud capex (accounts for as much as 12% of revenue, higher percent of profits), we lower our FY16 (fiscal year-end June 2016) revenue and EPS to $13.3B and $6.35, from $14.6B and $8.66 previously. Our estimates are now well below consensus (by 6% for Revenue and 16% for EPS). Additionally, slower unit growth in the cloud category, combined with lower client and enterprise units, will pressure fixed cost absorption. Our lowered estimates still assume WD can deliver gross margin in the bottom half of its 27-32% gross margin target, but weaker volume and enterprise/cloud mix could push it lower in the absence of meaningful flash storage growth or additional cost cutting opportunities. Lower PT to $70 from $113: Our prior $113 PT credited WDC with multiple expansion from historical levels as cloud became a larger growth driver and inventory/pricing behavior reflected a more consolidated industry. We now believe that with weaker demand across PC, enterprise, and cloud that P/E is likely to return to levels closer to the three year average range of 8-10x. Our $70 PT is based on 11x our lowered FY16 EPS of $6.35 as compared to a 13x P/E implied by our prior target of $113. The new 11x P/E is a point higher than Seagate's to reflect optionality from MOFCOM decision related cost take-outs and better execution in flash and is toward the low-end of IT Hardware ranges, reflecting our view that top-line will decline by 9% in FY16 and EPS by 16% (more than revenue due to margin weakness from lower fixed cost absorption and fewer mix benefits). Where we could be wrong – bull/bear case view: Our bull case of $119 stock price assumes the weakness in cloud data center growth is more moderate, allowing WD to hit at least the midpoint of its 27-32% gross margin target range and that MOFCOM approval allows for additional reductions in operating expenses. These upside drivers allow EPS to grow 13% this fiscal year, ahead of the bull case EPS trends for Seagate due to more OpEx opportunities from consolidating HGST assets if MOFCOM approves. In our bull case we assume shares can reach recent peak P/E of 14x given double-digit earnings growth. Our bear case of $45 assumes gross margin at the low-end of the 27-32% target range as weaker volumes and lack of MOFCOM approval offsets improving mix from the emerging flash segment. 10x bear case EPS of $4.50 assumes P/E multiple returns to the high-end of WD’s three year average range of 8-10x and in-line with the lower end of ranges for IT Hardware peers. 28

Global Technology| July 22, 2015 WWDDCC RRiisskk RReewwaarrdd Weak enterprise and cloud increase estimate risk with SSD growth an offset IInnvveessttmmeenntt TThheessiiss $140 Slowing enterprise and cloud demand make us less 120 $119.00 (+51%) constructive despite recent product portfolio investments and cost cutting opportunities. 100 $78.69 KKeeyy DDeebbaatteess 80 $70.00 (-11%) Can Western Digital achieve the upper-end of its 60 targeted gross margin range in FY16? Not in the $45.00 (-43%) near-term given unit shortfalls. Longer-term, 32% 40 gross margin is possible given improving SSD and cloud mix despite weak volumes in other 20 segments. Further integration of HGST could add incremental cost improvements. 0 Jul-13 Jan-14 Jul-14 Jan-15 Jul-15 Jan-16 Jul-16 What is the long-term unit growth outlook? We Price Target (Jul-16) Historical Stock Performance Current Stock Price WARNINGDONOTEDIT_RRS4RL~WDC.O~ look for 10% unit decline this year followed by Source: Thomson Reuters, Morgan Stanley Research stabilization in FY17 as PC / enterprise declines slow on easier compares and cloud continues to Derived from the base-case scenario. Price Target $70 grow. WDC receives approval to fully integrate HGST, which lowers Bull $119 PPootteennttiiaall CCaattaallyyssttss OpEx by $400M annually. Gross margins move towards the mid- 14x FY16e EPS of $8.50 point of the company’s target range despite limited unit upside. Further accretion from HGST upon regulator WDC trades just below the market multiple but at the high-end of approval for full integration its three year average range given cost opportunities and a more SSD assets drive incremental growth competitive product portfolio post acquisitions. Better than expected PC or cloud unit growth Ramp of leading-edge products – high capacity Weak enterprise and cloud demand raise margin risk with Base $70 cloud and hybrid drives SSD traction a wildcard. While gross margin falls in FY16, we 11x FY16e EPS of $6.35 continue to model better performance than peer STX in light of Potential for increased cash return to shareholders higher SSD mix and better overall execution. WDC trades in-line Further unit shortfalls or increased channel with recent range, or 11x, which is slightly above the three year inventory levels could pressure margins further average of 9-10x in light of higher SSD exposure and MOFCOM Slowing growth in cloud or SSD could drive P/E optionality. multiple back to historical range Weak demand, lack of HGST approval, and few synergies Bear $45 RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett from recent acquisitions pressure margins. Gross margin falls 10x FY16e EPS of $4.50 toward the low-end of target range with enterprise/cloud declines SSD adoption in tablets and client PCs more than offsetting SSD growth. The stock trades back to the Higher market share targets at Toshiba high-end of its long-term P/E range prior to industry consolidation Weak PC demand and high inventory given the unit pressure and reduced margin profile. Slowing areal density / technology transition lowers margins Lack of approval for HGST integration Tax rate increases 29

Global Technology| July 22, 2015 SSeemmii RRiisskk RReewwaarrdd AAlltteerraa ((AALLTTRR,, EEqquuaall--wweeiigghhtt,, $$4400 PPTT)) Solid revenue performance, opportunity to improve margins, and grow non wireless infrastructure IInnvveessttmmeenntt TThheessiiss On March 26th, the WSJ reported that Intel was in talks to acquire Altera, though they were also clear that no deal had yet been finalized. Last week, press reports indicated that negotiations had been broken off. This potential transaction is the key driver of the stock from here, in our opinion. We expect the reacceleration in PLD revenues that began in 2014 to slow somewhat in 2015 as wireless infrastructure tailwinds abate; but this will lead to higher gross margins, more defensible revenue streams, and fewer multiple depressing concerns. After share loss in 2012 and 2013, and share gains in 2014, we expect slight gains for Altera in 2015 mostly due to Xilinx's last time buys rolling off (but Source: Thomson Reuters, Morgan Stanley Research also a bit more risk, as Altera's wireless infrastructure exposure is higher); but Altera's 14 Primary driver is potential for Intel to buy the company, given WSJ Price Target $40 nm products later this year could drive share gains. report that they are in talks, overriding the intrinsic valuations below; represents 16% premium to the stock price before that KKeeyy VVaalluuee DDrriivveerrss announcement, slightly below similar transactions (CSR (by Qualcomm) 23%, International Rectifier (by Infineon) 56%, Hittite Revenues driven by Telco/wireless (42% of revs), (by Analog Devices) 36%, and Spansion (by Cypress) 18%. This networking and computing (22%), industrial, valuation implies, 29x EPS for 2016, or 23x ex cash, and is above military, and auto (22%). Huawei and Ericsson are where Altera has historically traded but still a level that would >10% customers make a transaction accretive for Intel (see below). It's quite possible We find the case for FPGA revenue gains vs. that the premium would be higher given that factor, but that is overall logic compelling, but historical gains have offset by the possibility that the deal won't happen. been mild with a 6.3% 5 year CAGR (vs. 5.9% for logic); we do expect a 2015 rebound / catch-up Upside from share gains and better growth in Bull $48 communications end markets, with 15-20% growth next 2 ~20x C16e Bull Case EPS of PPootteennttiiaall CCaattaallyyssttss years, and multiple expansion from strong 14 nm data points $2 + tax adjusted cash which lead to multiple expansion, as investors extrapolate share Ramp in wireless deployments in China, US and gains. India Data points on 20-nm and FINFET competition Expect ~11% decline 2015 and 7% growth in 2016. After Base $40 with Xilinx as an indication of 2014/15 share growing 11.5% with the wireless surge in 2014, we look for slower ~17x C16e GAAP EPS of growth in the next two years, but positive mix shifts - away from $1.36 + tax adjusted cash , in RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett wireless infrastructure, and towards server acceleration, SDN line with historic Slower growth in communications infrastructure, systems, ADAS, and other high growth markets. 14 nm is not a with highest variance from China spending game changer but leads to some unique TAM expansion opportunities. Competitor Xilinx 28-nm roadmap has several innovative new aspects with stacked silicon interconnect, integrated ARM/FPGA, and had very Downside from slow communications growth and a Bear $22 high market share through 2013, which reversed significant inventory correction. Large inventory correction ~13x C16e Bear Case EPS of in 2014. coupled with sluggish demand affects revenue growth. Assumes $1.10+ tax adjusted cash 11% revenue decline in 2015 with 67% gross margins, and flattish Company’s gross margin forecasts are more bullish revenue and 67% gross margin in 2016. than ours 30

Global Technology| July 22, 2015 AAvvaaggoo TTeecchhnnoollooggiieess ((AAVVGGOO,, OOWW,, $$116600 PPTT)) We like AVGO for its above-average growth, margin expansion, and M&A optionality - Overweight IInnvveessttmmeenntt TThheessiiss We are Overweight AVGO as the company’s above-average growth, strong FCF, diversified revenue stream (after LSI), and consistently high profitability is not reflected in the current multiple of P/E of 13.7X on CY15e EPS,~12% discount to S&P 500. We remain bullish on growth in the company's wireless business, which should drive upside to Street estimates. We also see optionality in the stock as Avago creates value through M&A. PPootteennttiiaall CCaattaallyyssttss Rising penetration of LTE handsets to drive higher dollar content (increase of 2-3X vs. 3G) and upside to estimates Source: Thomson Reuters, Morgan Stanley Research Upside to growth in comm for both ASICs and optical We value AVGO at ~17x CY2016e EPS of $9.25. We view this Price Target $160 multiple as conservative relative to analog peers/RF trading at 17- 20X and the S&P 500 (17.5x), given stronger earnings growth for Avago. RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett Stronger than expected competition or margin Revenue growth surprises to the upside and Avago exceeds Bull $190 pressure in FBAR, an important growth driver its opex savings target. The stock’s valuation multiple 19x 2015e ModelWare EPS of Execution on the LSI acquisition expands as investors gain confidence in the combined $9.99 company, including the merits of increased diversification and stability. – 13.7% rev growth in 2016. – GM increases 280bp Y/Y to 61.1% in 2016 Avago posts 10% revenue growth in CY16 on continued Base $160 strength in wireless. Wired infrastructure grows HSD and ~17x 2015e ModelWare EPS Enterprise Storage is flat to up LSD. Opex savings are in line of $9.25 with management guidance. – 9.4% revenue growth in 2016 – GM up 230 bp Y/Y to 60.6% in 2016 – ~$190mn in opex savings in CY15 relative to management’s target of a $200mn run rate exiting FY2015 – CY15 and CY16 EPS of $8.26 and $9.41 Revenue growth disappoints and management fails to realize Bear $109 targeted opex cuts. The stock’s valuation multiple falls to 14.5x 2015e ModelWare EPS 14.5x. of $7.53 – Revenue is up only 2.8% in 2016; GM up 80 bp to 59.1% 31

Global Technology| July 22, 2015 BBrrooaaddccoomm ((BBRRCCMM,, EEqquuaall--wweeiigghhtt,, $$5588 PPTT)) Overweight as steady, profitable, sustainable growth with market share gains drives multiple expansion IInnvveessttmmeenntt TThheessiiss Stock price mostly driven by the Avago bid, for a deal that we see as highly accretive for Broadcom, and only slightly above our previous 12 month price target. #1 position in high margin Broadband and Infrastructure markets should drive market share growth and create further separation from its competitors; Broadcom has a 40% share in Communication IC market with the number 2 competitor at < 10% Re-focusing on the fragmented ex-wireless communication IC markets should enable Broadcom to gain market share and target adjacent opportunities that had previously been ignored Source: Thomson Reuters, Morgan Stanley Research KKeeyy VVaalluuee DDrriivveerrss Our price target assumes 15.5x 2016 Non-GAAP EPS of $3.19 Price Target $58 (includes stock based compensation) plus nearly $8 in tax adjusted Mid-single digit revenue growth with lower net cash. This is roughly equivalent to Avago's bid for the company operating expenses should lead to high assuming some upside for Avago stock (with the buyout for both incremental margins and mid-teens EPS growth cash and stock). High semis R&D creates a competitive moat around the highest market share semiconductor Broadcom receives a superior offer compared to current Bull $67 companies leading to high ROIC and positive EVA ~$37bn transaction with Avago 18.5x 2016 EPS of $3.19 plus returns cash PPootteennttiiaall CCaattaallyyssttss Fundamentally, maintain dominant share in flagship Base $58 Continued enterprise recovery in 2015 and connectivity, LTE wind down as advertised, slow growth in 15.5x 2016 EPS of $3.19 plus continued standardization of networking around Networking with share gains at the expense of ASICs: mid ~$8 tax adjusted net cash standard, highly integrated chips should lead to single digit organic growth (ex baseband exit) in 2015 after modest strong revenue growth in networking Y/Y revenue growth in 2013/14. We model that Broadcom hangs Strong uptake of Cisco Nexus 9000 switches on to most connectivity platforms for high end flagship phones, with some pressure in the midrange. Stock price similar to intrinsic Market share gains tied to secular trends like the value, but mostly driven by the Avago purchase price. growth of merchant silicon due to SDN overlays in networking; connectivity could be a core asset in Deal with Avago falls through because of fundamental greenfield opportunities like IOT and wearables Bear $35 erosion. Also, the company loses share in flagship ~12x 2016 EPS of $2.86 plus Flagship phones from Apple and phones and connectivity (>35% of revs and 25% of earnings) in 2016 and net cash tablets from Samsung will prove out the the networking Trident 2 upgrade cycle runs out of steam. company’s flagship connectivity share claims We view this scenario as unlikely, given Avago's commitment. RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett Loss of flagship handset wins would impact connectivity (20% of earnings) Networking rebound somewhat macro-economic- dependent; base station business has minimal growth 32

Global Technology| July 22, 2015 CCaavviiuumm ((CCAAVVMM,, OOvveerrwweeiigghhtt,, $$6688 PPTT)) High Growth Justifies High Valuation IInnvveessttmmeenntt TThheessiiss Cavium should maintain share in the highest growth portion of the communications processor market (>8 cores), thus outgrowing the peer group There are a number of growth initiatives that take time to play out, but allow for optionality in the 2015-17 time frame All of this growth comes at a price, as the stock trades at a 30%+ premium to comm IC peers on a relatively low quality non-GAAP earnings number SShhoorrtt--tteerrmm CCaattaallyyssttss Comm infrastructure environment strengthening, particularly service provider/wireless with China Mobile ramp Source: Thomson Reuters, Morgan Stanley Research Several growth headwinds starting to fade (exit of Equal to our Base Case scenario. Our Price Target of $68 is equal Price Target $68 software business, flattish Cisco because of delays to 26x 2015e Non-GAAP EPS of $2.44 (slightly above historical with Cat 3K) multiples given upside from ARM server opportunity) New business (LiquidIO) starting to ship into a large hyperscale datacenter; drives our revenue – Core business continues to execute in line with the base case, Bull $92 upside vs. consensus next year basically holding share in the high end of the communications 28x bull case 2016e EPS of processor business in a strengthening environment for comms $3.30 LLoonngg--tteerrmm CCaattaallyyssttss – But the newer initiatives, particularly Project Thunder, start to materially show that Cavium will be one of the relevant 2-3 Project Thunder (microserver) market entry highly vendors in the early stage of ramping high density server CPUs, uncertain but could be a large opportunity driving $100 mm+ revenue upside in 2016 (& multiple expansion Multiple other growth initiatives within small-cell given the $14 bn server CPU TAM) basestations, TCAM replacement, share gains in lower end – 14.8% revenue growth in 2015, a deceleration based on slower Base $68 wireless infrastructure; growth driven by content growth of higher- 26x non-GAAP 2016e EPS of RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett end communications processors, and small contribution from new $2.44 + cash product initiatives next year Valuation highest in the group, leaves little room –Expect acceleration to 27% growth next year, as Thunder for error (microservers), xPliant (switches), and liquid I/O 2 all significantly Customer concentration risk (Cisco 18% of revs, expand the TAM; we see Cavium as a niche player in all three top 3 customers 51%) markets but still can add 10-20% to growth rates Higher capitalized competition focused on this market (Broadcom, Intel, Freescale). – Communications infrastructure markets sputter again as they did Bear $40 in 2012, with deceleration in both service provider and enterprise 22x 2016e bear case EPS of – Timing of several key growth areas continues to be pushed out; 1.60 microcell basestations, TCAM replacements; Cavium mostly hits product milestones but design-ins take longer – Project Thunder server timing takes longer, depressing earnings but maintaining a premium valuation 33

Global Technology| July 22, 2015 IInnpphhii ((IIPPHHII,, EEqquuaall--wweeiigghhtt,, PPTT $$1188..55)) iMB and 100G Product Cycles Drive Upside IInnvveessttmmeenntt TThheessiiss The acquisition of Cortina, which closed in 4q14, should add substantially to 2015 earnings, by nearly a factor of 2, since Cortina is highly profitable and doesn’t have the high investment of the core business. With most of its revenues exposed to secular growth markets in Data Center/Cloud (~60% of revenues) and Data/Video transport (~40% of revenues), we model Inphi’s ex-Cortina top-line grows 27% in 2015 and 15% in 2016. The company’s core strength in high-speed analog design is driving its products for data centers and telecom, with strong growth potential over the next three years. Communications has just outgrown the Server Source: Thomson Reuters, Morgan Stanley Research segment – driven by 100G products and continues to be the main growth driver. Equal to Base Case scenario. We value Inphi on 15x 2016e EPS , in Price Target $18.5 line with the peer group of infrastructure exposed names (BRCM, KKeeyy VVaalluuee DDrriivveerrss CAVM, ALTR, XLNX, MRVL) Growth in worldwide Datacenter and Cloud Cortina acts as an accelerant to communication segment growth, Bull $27 infrastructure builds and IPHI’s iMB product ramps faster than expected, maintaining a 16.5x Bull Case CY16 non- Ramp of iMB on INTC Haswell server platform with higher share of the market than we are modeling. IPHI quickly GAAP EPS of $1.63 increasing DDR4 LRDIMM attach rates recovers market share in DDR4 buffers despite it's late start. Continued worldwide build out of 100G Strong market growth and market share gains at 100G. infrastructure Capex/Opex savings through the use of coherent Cortina doesn’t grow much but substantially contributes to higher Base $18.5 optics drive upgrade of optical modules EPS. IPHI’s iMB product attach rate improves in 2015-16, 100G 15x Base Case CY16 non- starts gaining traction. Attach rates continue to move up in 2015. GAAP EPS of $1.27, in line Growth in 100G continues in 2015 and into 2016. Top line growth PPootteennttiiaall CCaattaallyyssttss with infrastructure related of 15%+ post-Cortina but continued aggressive investment for peer group Further launches of server series using LRDIMMs longer-term growth caps near-term EPS upside (e.g. HP Proliant) and industry studies highlighting the benefits of LRDIMMs Traction with Intel’s Haswell platform more drawn out with muted Bear $11 Price declines of LRDIMM modules to <$1000 per adoption of iMB pushing out IPHI’s revenue ramp. Competitive 8.5x Base Case CY16 adj. EPS module landscape toughens, adding price pressure for LRDIMMs in DDR3 of $0.92+~$3 in cash Capex announcements from carriers for 40G/100G and loss of market share in DDR4. 100G ramp is slower than ports and switches expected. Corresponding lack of visibility drives multiple compression RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett Lower attach rates of DDR4 LRDIMM modules and loss of market share A stall in the worldwide build out of 40G/100G comms infrastructure Unfavorable outcome of litigation with Netlist 34

Global Technology| July 22, 2015 IInntteell ((IINNTTCC,, UUnnddeerrwweeiigghhtt,, $$3311 PPTT)) Inventory grows to new high with 10nm Cannonlake pushed to 2H17 acting as a tailwind to FY15 margins, but aggressive growth targets for 2H15 makes it difficult for us to be constructive IInnvveessttmmeenntt TThheessiiss Core growth in PCs and servers is still key growth driver; expect minimal growth in 2016 Non PC/server initiatives likely continue to disappoint Foundry customers still unclear, beyond Altera, Panasonic which are small; we think building a sustainable earnings stream from foundry will be challenging. Tablet subsidies in 2014 makes sense, in that it builds critical mass around Android on x86. But turning it profitable will be a challenge. Delays in SoFIA LTE does not help with cost savings in 2015. We are reasonably upbeat on prospects for servers but view 15% y/y growth as too high Process node transitions taking longer; the Source: Thomson Reuters, Morgan Stanley Research company’s 2 year “tick/tock” cadence is clearly Base case scenario. Price Target $31 under stress with the launch of a 3rd year tock variant in Kaby Lake. PC market accelerates in 2H15 and 2016, Data center Bull $43 remains robust, and Intel’s success in PPootteennttiiaall CCaattaallyyssttss 17x 2016e EPS of $2.55 tablets/phones/foundries drives multiple expansion Skylake launch in 2H15 is insufficient to drive PC – While investor focus is on foundry and smartphones, the key recovery with Win10 being less of an event for earnings driver remains PC and servers new purchases – Data center growth accelerates to 20% growth, as PCs grow Grantley server platforms in 2H adoption slows as modestly DCG misses its 15% y/y target growth rate – Lower capex could fill the free cash flow/earnings gap Channel inventories remain elevated resulting in sharp deceleration in utilisation levels PC correction has largely played out with y/y growth starting Base $31 to improve but still see inventory levels as potentially too ~14x 2016e EPS of $2.20, high and revenue risk in 2H15 due to aggressive RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett slight discount to the group expectations; we project: (in line with history) Biggest upside in our view would come if Skylake – Gross Margins: 61.8% GMs in 2015e and 62.5% in 2016e performance drives uptick in PC demand in 2H15 – 2015e and 2016e EPS of $2.12 and $2.20, respectively 3rd year Tock variant in Kaby Lake could drive – Mixed success in key non x86 initiatives (foundry, smartphones) margin upside with lower depreciation, more mature yields and lower raw material cost Assumes average selling price and units both decline 5% Bear $20 2015 and 2016 due to mix shift to Bay Trail notebooks, 14x 2016e EPS of $1.43 without offsetting volume, and servers slow dramatically – Revenues declines at 4% 2015 – Return on assets declines from 18% to 12% by year end – Dividend perceived to be at risk – Multiple compression as concern builds about further margin degradation in 2016 35

Global Technology| July 22, 2015 MMaarrvveellll TTeecchhnnoollooggyy ((MMRRVVLL,, OOvveerrwweeiigghhtt,, $$1177..5500 PPTT)) Risk Reward Snapshot: Marvell Technology (MRVL, Overweight, PT $17.50) IInnvveessttmmeenntt TThheessiiss We expect the company to sustain its market share in storage and grow revenues from rising demand for storage capacity despite the declining PC market. With an 18% FCF margin and a share count that’s declined by over 130m shares to 530m over last 2 years through share repurchases, the stock offers long term value, but with near term pressures in SSD and baseband markets PPootteennttiiaall CCaattaallyyssttss Carnegie Mellon litigation: Oral arguments for the Marvell appeal next couple of months, resolution later this year; we expect a sizable fine but would be surprised if there are ongoing royalties Source: Thomson Reuters, Morgan Stanley Research Marvell has grabbed 2nd place in China TD LTE; Midway between base and bull case, given likelihood of some Price Target $17.50 how sustainable is that? The company has argued resolution of baseband in the next 12 months for a much stronger portfolio vs. the TD-SCDMA portfolio a couple of years ago, which fell flat after – Only difference from the base case is potential for baseband the initial surge Bull $21 loses to go away - $348 mm revenue at 38% gross margins drives 14x CY16e MW ests ex Additional wins in LTE outside China $132 mm of gross profit, funding $400 mm in R&D and $50 mm baseband losses of $1.27, + in SG&A, which drives a $318 mm pretax loss; leads to $1.22 MW $3.5 cash ex CMU RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett EPS CMU litigation: Marvell was found guilty of violating Carnegie Mellon IP, resulting in a $1.17 – We expect storage to continue to rebound after a tough 1h15 Base $14.50 bn fine and a $0.50 per hard drive unit royalty. We – Mobile and wireless declines in CY15, then grows somewhat in $14.50 (16x CY16e MW EPS expect damages could be reduced on appeal CY16, on the back of LTE, but remains subscale and unprofitable of $0.67, plus adjusted cash, – Relatively stable networking revenues The strategic nature of baseband is driving in line with the group and a – Long term we think baseband improves, as either the company excessive investment, and we don’t foresee returns premium to historic given the grows towards critical mass or reduces investment. We also see the for anyone other than the leaders QCOM and potential to fix baseband elimination of the Carnegie Mellon litigation as a source of upside. Mediatek. losses) – 2 major differences from the base case: 1) CMU litigation ends Bear $11 up with damaging royalties, which reduces operating profit per $11 (14x depressed CY16e unit in storage by about 60% bear case EPS – 2) Worst case scenario in Mobile and Wireless where the company loses LTE traction and accelerates R&D 36

Global Technology| July 22, 2015 MMiiccrroonn ((MMUU,, EEqquuaall--wweeiigghhtt,, $$1199 PPTT)) Negative on fundamentals, but potential Tsinghua interest limits downside IInnvveessttmmeenntt TThheessiiss We expect PC DRAM to remain weak in 3Q, different from our previous assumption of a mobile lead recovery. We do expect stabilization in calendar 4Q, but see supply growing slightly faster, and demand growing slightly slower, beyond that time. We continue to believe that memory stocks have a relatively well defined earnings cycle,though highs and lows are likely to be higher than they have been historically; SShhoorrtt--tteerrmm CCaattaallyysstt Pricing remains firm in mobile, offsetting the weakness in PC Source: Thomson Reuters, Morgan Stanley Research LLoonngg--tteerrmm CCaattaallyyssttss Base case scenario. Price Target $19 We think there is likely more DRAM supply than consensus believes in late 2015 and in 2016. Strong memory cycle returns next year. DRAM goes back into Bull $30 Consensus is for a sharp slowing in supply from allocation mode in late 2015, around stronger content in mobile 10x $3 bull case CY16e EPS, 2014 levels, yet capital spending is much higher and a seasonal pickup in Qcs, and then settles back down (but still and investor belief that it is (even considering capital intensity issues) and the at high levels) in 2016. NAND is slightly better than our base case normalized earnings level; Hynix fire and Micron fab conversion are non scenario, which is already bullish; Micron achieves new highs in recurring. gross margin by 1Q15. We still see 2015 and 2016 as good earnings years for the industry, just not as good as the period Company behavioral changes make investors believe there following a large supply shock to the industry in are sustainable profits to be had here; companies leave fab 2014. utilization below 100% even during relatively profitable periods, Micron has some optionality if they can improve slow capacity shrinks further, make no capacity announcements. profitability in NAND, which has been below peers. We'd become more constructive if we saw companies behaving differently reflecting a more consolidated industry, but so far they have not. RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett Pricing can turn quickly; while our channel contacts After prices soften slightly through 2015, we have a “normal” Base $19 are largely negative in the short term, a surge in year with 20% price declines in 2016. While this seems 11x 2016e EPS of $1.76, in demand could turn that around quickly. We think draconian relative to the last two years, we note that $1.76 would line with historic periods of it's more likely to surprise in the other direction, be one of the best earnings year in history before the 2013-15 strength given elevated inventory levels. period, and still reflects a much better industry dynamic than we Longer term, we are watching Samsung's spending have seen historically. We see mild pockets of oversupply in 2016, patterns closely. They outgrew industry supply not a return to historic downturns. (and their forecast) substantially in the last 2 years, but now are aware that Micron and Hynix are likely to see accelerated supply. Will they willingly lose share to preserve profits? 37

Global Technology| July 22, 2015 $16 bear case in part to account for Tsinghua Unigroup's bid for Bear $16 the company as reported by WSJ as we anticipate greater focus on 1.6x tangible book value, with Micron's potential strategic value. downside limited by potential strategic interest Overall in our bear case, despite slow supply growth, DRAM prices collapse, in a scenario very similar to 2012 (when record slow supply growth did not lead to good returns). Modest weakness in NAND in midyear, worsened by capacity announcements that create 2014 concerns. Elpida assimilation proves more difficult than expected, as the 20 nm migration requires more spending, and elongated technology transitions cause earnings weakness similar to what we saw during the Inotera integration. SSaannDDiisskk ((SSNNDDKK,, EEqquuaall--wweeiigghhtt,, $$7755 PPTT)) Improved NAND Market is offset by execution challenges IInnvveessttmmeenntt TThheessiiss SNDK is a beneficiary of the slowing of Moore’s law, with lower capital intensity and a low-cost leadership position in 2D planar technology. 3D remains a debate but generally we think the slower supply effect of 3D investment is more important then product characteristics Demand accelerates as NAND manufacturers capture a larger share of consumer’s willingness to pay for higher NAND content through lower mark- ups on NAND in smartphones and tablets Equipment orders from NAND are shrinking, and the 3D investment should slow incremental supply per $ of new capacity Sandisk has experienced significant challenges in monetizing a healthy NAND environment, which Source: Thomson Reuters, Morgan Stanley Research we attribute to an excessively vertically integrated enterprise strategy; further, we think the heavy Base case scenario. Price Target $75 investment in enterprise has distracted the company's execution in its core business. – NAND supply remains constrained, and drives EPS to $6 in 2016 Bull $99 – Prices rise in 2h15, in conjunction with yen based cost declines, ~15x our bull case CY16e EPS PPootteennttiiaall CCaattaallyyssttss driving GMs to high 40s of $6, + net cash – Enterprise SSD grows to the company’s target of $1bn, driving Sandisk struggling with several self inflicted continued higher valuation problems; we think they could recover relatively – Elasticity at work in smartphone/tablet segments as OEMs cut quickly with a good supply/demand backdrop. price on >32 GB upgrades Recent pricing has been benign despite weaker – SanDisk adds sustainability to business model with higher demand; we are optimistic about 2h pricing. enterprise SSD concentration, & builds an enterprise storage business, driving higher multiple; SSDs make up over a third of RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett revenue Demand weakness in basically all NAND consuming markets in 1Q (PC, smartphone ex – Sandisk makes some progress fixing company specific issues by Base $75 apple, retail) 2h15; still below model GMs in 2h and into 2016 13x CY16e EPS of $5.06 plus – Improved supply demand environment in 2h and into next year, Limited visibility into SNDK specific problems due $8.70 net cash, consistent as capital spending has declined and demand is elastically to poor management disclosure with typical mid cycle memory accelerating multiples – We do assume retail rebounds somewhat in 2h; Assumes limited progress in enterprise in 2015, more so in 2016 38

Global Technology| July 22, 2015 – NAND oversupply recurs and earnings remain sub $3 Bear $54 – 3D NAND supply is more viable than we expect, and ramps faster Derived from 70% of – Asset values protect the stock. We lower royalties DCF to $10 in replacement cost ($12.5bn) + an oversupply situation. cash – These scenarios seem far fetched from 30% operating margins, but if prices fall 15% more than expected it costs $3 in EPS XXiilliinnxx ((XXLLNNXX,, EEqquuaall--wweeiigghhtt,, $$4433 PPTT)) We expect sales to decline further in FY16 before recovering FY17; Intel buyout of Altera raises questions about profitability IInnvveessttmmeenntt TThheessiiss On June 1st, Intel announced that it would acquire Xilinx's largest competitor, Altera. This creates potential concerns that Intel might become more aggressive on pricing, using its incremental foundry margin to try to drive share gains on next generation technologies. In the middle of a large, though lumpy, LTE deployment which should drive growth next 2 years 28nm product portfolio seems innovative and has high potential, though really a return to normal after Altera had very strong 40 nm share Next battleground at 20 nm + 16 nm vs. Altera/Intel at 20 nm+14 nm Source: Thomson Reuters, Morgan Stanley Research KKeeyy VVaalluuee DDrriivveerrss Base case scenario. Price Target $43 Revenues driven by Industrial (~39%), consumer/auto (~15%), and other (~3%) Given that Altera was taken over by Intel at roughly 34x EPS + Bull $60 We see continued share gains, as the company cash, the market could assign a higher valuation to Xilinx as the 21x 2016 EPS of $2.54 + should show high 20 nm share as it rolls out, for 2 only standalone FPGA vendor reflecting the strategic nature of cash, based on strategic value successive nodes of share gains FPGAs. If Intel is successful at driving revenue synergies, the Xilinx We find the case for FPGA revenue gains vs. business could be viewed as increasingly attractive to companies overall logic compelling, particularly as seeking a presence in the data center. communications, enterprise, and industrial markets are relatively appealing vs. Similar to current valuation, as growth concerns persist. Sales Base $43 consumer/smartphone/PCs decline 1.7% in FY16 with Communications & Data Center down 14x Base Case CY16e EPS of 10%, Industrial, Aerospace & Defense up 5.1% and Broadcast, $2.54, plus $7.5 in cash PPootteennttiiaall CCaattaallyyssttss Consumer & Auto up 7.1%. Our target multiple is slightly below the midpoint of the long term multiple range that the stock has Upside in wireless revenues if China FD LTE orders traded at over the last five years ramp earlier than expected and LTE sending in other geographies such as Europe and India picks Weak end markets, reversal at 14 nm as Altera takes leadership up Bear $32 share, no leverage. Stall in communications infrastructure end- 12x Bear Case CY16e EPS of Continued ramp in 28nm revenues, as a precursor market growth results in a revenue. Further decline 2016. $2.02, plus $7.5 in cash to 2015/16 market share RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett China spending could be lumpy Altera’s move to Intel as a foundry could create a process disadvantage for Xilinx if it regains leadership at 14nm FINFET, though revenue impact is largely 2016 and beyond ALTR’s 40nm share gains are still a small headwind 39

Global Technology| July 22, 2015 40

Global Technology| July 22, 2015 CCSSCCOO RRiisskk RReewwaarrdd Concerns over SDN/NFV technology disruption already priced in IInnvveessttmmeenntt TThheessiiss The market is still recovering from being overly concerned about the near-term adverse impact of SDN and NFV on Cisco’s business model. Cisco's earnings and growth rates are largely tied to IT spending, which is closely correlated to GDP growth. We expect that new products, particularly in the switching and routing segment, are likely to take small share in 2014 and 2015, which, because of Cisco's size and scale, we anticipate will be a primary growth driver for the company. If the company fails to hit our forecasts, Cisco management could be motivated to meaningfully reduce operating expenses. Source: Thomson Reuters, Morgan Stanley Research estimates KKeeyy DDeebbaatteess 13x FY 16e EPS Price Target $30 Does SDN and NFV eviscerate CSCO's traditional business model? We think customers are Share gains across multiple units, operating margin prioritizing OPEX reductions versus CAPEX Bull $36 improvement . Cisco is able to generate share gains across reductions as the goal of SDN, reducing upfront 15x FY 16e EPS multiple sectors, resulting in revenue growth that is in the 7-9% pricing pressure. range. At the same time the company reduces R&D as a Does the move to the cloud adversely impact percentage of revenue, resulting in ~100 bps of operating margin CSCO's traditional enterprise market? Yes, but improvement per year for the next 2-3 years. Better-than- over a long term, in our view. expected results would likely lead to modest multiple expansion to Could we see meaningful multiple expansion on 15x earnings. further reduction in SDN- and NFV-related concerns? Yes, we believe it has contributed to Share gains in switching, operating margins stay steady. Cisco Base $30 recent expansion and will contribute more. is able to generate share gains with its new switching products 13x FY 16e EPS (Cisco’s largest segment, 30% of revenue), while maintaining share PPootteennttiiaall CCaattaallyyssttss in most other segments, leading to annual revenue growth of 4- 6% with little to no operating margin leverage expansion. On any Better than expected uptake on new switching move to actually reduce operating expenses (i.e., increase revenue products. per employee and/or reduce R&D as percentage of revenue), we Aggressive operating expense cuts implemented would expect multiple expansion to 13x out-year EPS. without suffering through disappointing results. An acceleration in GDP and therefore IT spending. Switching growth flattens, operating margins compress as Bear $19 A reduction in semiconductor CAPEX. SDN and NFV are adopted faster than expected. Increased 8x FY 16e EPS compute power applied to networking accelerates SDN and NFV, leading to flat revenue growth and margin compression of 100- RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett 200 bps per year. Under such a scenario, we would expect the Slower than expected GDP and IT spending multiple to contract to ~8x out-year EPS, with downside protection growth. being afforded by the large cash balance, dividend (currently ~3% Failure to take sufficient share with new switching yield) and buy backs (currently authorized for $12.1 billion). and routing products. New architectures (e.g., SDN and NFV) subject Cisco to more pricing and margin pressure than currently anticipated. 41

Global Technology| July 22, 2015 AAssppeeeedd EEaarrnniinnggss RReevviissiioonn && VVaalluuaattiioonn MMeetthhooddoollooggyy We are revising down our 2015 revenue forecast by 3% given server demand postponement in 2Q15. We also lower our 2016/2017 revenue forecast slightly by 4%/ 4%. As a result, our 2015-2017 EPS estimates are coming down by 4%/ 5%/ 3%, respectively. EExxhhiibbiitt 1166:: Aspeed: Earnings Revisions Current Previous Current Previous Current Previous NT$ mn 2015E 2015E Diff. 2016E 2016E Diff. 2017E 2017E Diff. Net sales 1,024 1,053 -3% 1,480 1,534 -4% 1,799 1,872 -4% COGS 457 470 647 671 785 817 Gross profit 568 583 -3% 833 863 -4% 1,014 1,055 -4% Operating expenses 188 188 285 286 349 365 Operating profit 380 395 -4% 549 578 -5% 665 690 -4% Non-op. income (exp.) 5 5 8 8 8 8 Pretax Income 385 400 -4% 557 586 -5% 673 698 -4% Taxes 46 48 77 82 93 97 Net income 339 352 -4% 479 504 -5% 580 601 -3% Reported EPS 12.93 13.45 -4% 18.29 19.23 -5% 22.14 22.93 -3% Margins Gross margin 55.4% 55.4% 0% 56.3% 56.3% 0% 56.4% 56.4% 0% Operating margin 37.1% 37.5% 0% 37.1% 37.7% -1% 37.0% 36.9% 0% Pretax margin 37.6% 38.0% 0% 37.6% 38.2% -1% 37.4% 37.3% 0% Net margin 33.1% 33.5% 0% 32.4% 32.8% 0% 32.2% 32.1% 0% Source: Morgan Stanley Research estimates We derive our new price target of NT$340 (down from NT$350 previously) from our residual income model. We assume a long-term growth rate of 5.2% (unchanged) and a cost of equity of 9.8% (2.0% risk-free rate, 6% risk premium, beta of 1.3). We lower mid-term growth rate from 7.2% to 6.8% in view of recent soft server demand. Our risk reward scenario values are changing slightly as a result of the changes to earnings. Our scenario analysis suggests that Aspeed's risk-reward is fair. Bull-case value – NT$475, implying 37x our 2015E EPS forecast: Aspeed’s product performance outweighs peers’ and its gross margin improves significantly after the migration to 40nm. Earnings post robust growth on booming data center demand thanks to rising enterprise cloud adoption and government support, with white- brand servers dominating the server market. Aspeed enjoys robust TAM growth and gains market share in brand server names like Dell. Virtual desktop business takes off with enterprises accelerating virtual desktop adoption. Base-case value – NT$340, implying 26x our 2015E EPS forecast: Aspeed’s gross margin holds up well, given its competitive cost structure after technology migration. Revenue expands as data center demand enlarges its white-brand server exposure, and it gains share at Supermicro. Virtual desktop gradually ramps with a noticeable installation base increase due to the need for lower maintenance cost and more flexibility. Bear-case value – NT$171, implying 13x our 2015E EPS forecast: Gross margin pressure emerges as peers’ technology capability and migration speed up. Enterprise IT spending slows down on macro conditions, capping data center market growth, and competition intensifies accordingly, given the limited growth of the total addressable market. Virtual desktop development remains slow; Microsoft, Aspeed’s key partner, has difficulties penetrating the market. 42

Global Technology| July 22, 2015 EExxhhiibbiitt 1177:: Aspeed: Residual income model NT$million 2015E 2016E 2017E 2018E 2019E 2020E 2021E 2022E 2023E 2024E 2025E 2016E Total Equity 1,015 1,239 1,435 1,636 1,851 2,081 2,326 2,588 2,868 3,167 3,486 3,486 Net Profit 339 479 580 620 662 707 755 806 861 919 982 982 ROAE 36.1% 42.5% 43.4% 40.3% 37.9% 35.9% 34.3% 32.8% 31.6% 30.5% 29.5% 29.1% Residual Income 227 332 416 438 461 484 509 535 563 593 624 610 Spread 26.3% 32.7% 33.6% 30.5% 28.1% 26.1% 24.5% 23.0% 21.8% 20.7% 19.7% 19.3% Ending Equity Capital 1,015 PV of Forecast Period 2,935 PV of Continuing Value 4,964 Equity Value 8,914 No. of Shares 26 Projected Price 340 Source: Morgan Stanley Research estimates EExxhhiibbiitt 1188:: Aspeed: PE chart 24 90% ) 22 % ( h t w 20 45% o r g Y / 18 ) Y x ( e u 0% E / n 16 P e v e R y 14 l r -45% e t r a 12 u Q 10 -90% M A N F M A N F M e e u u o o a a a b b g g v v y y y - - - - - - - 1 - 1 - 1 1 1 1 1 1 1 4 5 3 4 3 4 3 4 5 P/E Multiple Average One Standard Deviation One Standard Deviation Revenue Y/Y Source: Thomson Reuters, Morgan Stanley Research 43

Global Technology| July 22, 2015 AAssppeeeedd RRiisskk RReewwaarrdd Balanced Risk Reward IInnvveessttmmeenntt TThheessiiss Aspeed should see margin expansion thanks to its competitive cost structure due to leading technology capabilities. The company looks well positioned to benefit from data center and growing white-brand server demand. PC/AV extension and virtual desktop are likely to become its next growth drivers if executed well. However, cloud service capex is decelerating, and our analysis suggests that the demand shortfall could be structural. We also believe that valuation is now fair. KKeeyy VVaalluuee DDrriivveerrss Source: Thomson Reuters, Morgan Stanley Research Market share in BMC Margins Residual income model. Price Target NT$340 New business progress Booming data center demand and further share gains; Bull NT$475 meaningful margin expansion; virtual desktop business PPootteennttiiaall CCaattaallyyssttss 37x 2015 EPS taking off: Significant earnings growth on booming data center Further market share gains at brand server demand and market share gains at both white-brands and other companies, such as Dell. brands. Margins improve considerably. Virtual desktop business Grantley platform rollout, speeding up as DDR4 growth accelerates. and CPU price drops to the sweet spot with sufficient supply. Strong albeit decelerating data center demand; mild margin Base NT$340 New virtual desktop IC design wins for Microsoft. improvement; virtual desktop business gaining traction: 26x 2015 EPS Revenue momentum strong with 30% CAGR in 2015-2016, thanks to both enlarged end-market share and new design wins. Gross RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett margin improves mildly. Virtual desktop business starts to bear Downside Risks: fruit. Weaker-than-expected data center demand and slower market share shifts toward white server Disappointing end-demand with share loss; margin pressure; Bear NT$171 names. virtual desktop business problems: Revenue growth decelerates 13x 2015 EPS Gross margin pressure on intensified competition. due to slow data center demand and share loss amid a weakening macro environment. Gross margin pressure emerges on intense Problems in virtual desktop development. competition. Virtual desktop business stays flattish, given slow Volatile project-based revenue. penetration and customers’ limited footprint. Upside Risks: Gross margin improvement, dependent on process node migration and cost-down efforts. Robust data center growth, mainly driven by both traditional OEM and cloud service provider demand. Market share gains at not only white-brand server makers like Supermicro but also brand server firms like Dell. 44

Global Technology| July 22, 2015 AAssppeeeedd:: QQuuaarrtteerrllyy aanndd AAnnnnuuaall EEaarrnniinnggss EEssttiimmaattee EExxhhiibbiitt 1199:: Aspeed: Quarterly and annual earnings estimates NT$ in million 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15E 3Q15E 4Q15E 2013 2014 2015E 2016E 2017E 0 0 0 0 0 0 0 0 0 0 0 0 0 Total Revenues 178 229 212 237 242 206 271 304 591 856 1,024 1,480 1,799 Sequential Change 20.2% 28.4% -7.1% 11.7% 2.2% -14.8% 31.2% 12.4% Change vs Year Ago 28.3% 46.7% 43.2% 60.2% 36.2% -9.7% 27.6% 28.4% 22.0% 44.9% 19.7% 44.5% 21.6% Cost of Sales 81 101 94 108 108 92 121 136 271 385 457 647 785 Percent of Revenues 46% 44% 44% 46% 45% 45% 45% 45% 46% 45% 45% 44% 44% Gross Margin 97 128 118 129 134 114 150 168 320 471 568 833 1,014 Percent of Revenues 54.2% 55.8% 55.7% 54.3% 55.5% 55.4% 55.4% 55.3% 54.2% 55.0% 55.4% 56.3% 56.4% Incremental Margin 63% 61% NM 43% 105% NM 55% 55% 48% 57% 57% 58% 57% Total Opex 35 44 43 44 42 43 51 52 138 166 188 285 349 Percent of Revenues 19.6% 19.3% 20.0% 18.7% 17.2% 20.7% 18.9% 17.1% 23.4% 19.4% 18.3% 19.2% 19.4% R&D 25 29 28 29 26 28 35 36 88 111 125 198 252 Percent of Revenues 14.1% 12.9% 13.1% 12.2% 10.8% 13.6% 12.9% 11.8% 14.9% 13.0% 12.2% 13.4% 14.0% General & administrative 9 9 10 10 10 10 10 10 35 38 40 54 57 Percent of Revenues 5.2% 3.8% 4.7% 4.2% 4.1% 4.8% 3.7% 3.3% 5.9% 4.4% 3.9% 3.6% 3.2% Selling & marketing 0 6 5 5 6 5 6 6 15 17 23 33 40 Percent of Revenues 0.2% 2.7% 2.3% 2.2% 2.3% 2.3% 2.3% 2.0% 2.6% 1.9% 2.2% 2.2% 2.2% Operating Income 62 83 76 85 93 72 99 116 182 305 380 549 665 Percent of Revenues 34.7% 36.5% 35.6% 35.6% 38.2% 34.7% 36.5% 38.2% 30.8% 35.7% 37.1% 37.1% 37.0% Change vs Year Ago 53.0% 83.7% 53.8% 79.2% 50.2% -14.1% 30.7% 37.7% 18.3% 67.7% 24% 44% 21% Total Non-operating Income(Loss) 3 (2) 7 11 (1) 2 2 2 12 20 5 8 8 Profit Before Taxes 65 82 83 96 92 74 101 118 194 325 385 557 673 Percent of Revenues 37% 36% 39% 40% 38% 36% 37% 39% 33% 38% 38% 38% 37% Taxes 7 12 10 15 11 11 11 13 23 45 46 77 93 Tax Rate 11.4% 15.3% 12.2% 15.2% 11.8% 15.0% 11.0% 11.0% 12.0% 13.7% 12.0% 13.9% 13.8% Net Income, Cont Ops 58 69 73 81 81 63 90 105 171 281 339 479 580 Percent of Revenues 32% 30% 34% 34% 33% 30% 33% 35% 28% 29% 33% 32% 32% Reported Income (TW GAAP) 58 69 73 81 81 63 90 105 171 281 339 479 580 Percent of Revenues 32% 30% 34% 34% 33% 30% 33% 35% 29% 33% 33% 32% 32% Change vs Year Ago 0% 0% 0% 0% 0% 0% 0% 0% 26% 64% 21% 41% 21% Reported EPS (NT$, TW GAAP) 2.42 2.91 2.77 3.10 3.09 2.40 3.43 4.02 7.39 10.72 12.93 18.29 22.14 Change vs Year Ago 28% 57% 47% 55% 27% -17% 24% 30% 14% 45% 21% 41% 21% Source: Morgan Stanley Research estimates 45

Global Technology| July 22, 2015 AAssppeeeedd FFiinnaanncciiaall SSuummmmaarryy EExxhhiibbiitt 2200:: Aspeed: Financial data Income Statement Cash Flow Statement NT$mn (Years End Dec ) 2014 2015E 2016E 2017E NT$mn (Years End Dec ) 2014 2015E 2016E 2017E 856 1,024 1,480 1,799 247 354 412 538 Net sales Cashflow from Operations (385) (457) (647) (785) 281 339 479 580 COGS Net profits 471 568 833 1,014 14 5 6 6 Gross profit Depreciation (166) (188) (285) (349) (73) 10 (73) (49) Operating expenses Working Capital Change 305 380 549 665 25 0 0 0 Operating income Other adjustments 20 5 8 8 (9) (10) (15) (18) Non-operating income Cashflow from Investing 325 385 557 673 (12) (10) (15) (18) Pre-tax income Capex 45 46 77 93 0 0 0 0 Income tax Change of LT Investment 281 339 479 580 0 0 0 0 Reported net Income Change of ST Investment 26 26 26 26 3 0 0 0 Adj.wtd.avg.shrs( m) Other adjustments 10.72 12.93 18.29 22.14 (86) (184) (285) (385) Reported EPS (NT$) Cashflow from financing 0 0 0 0 Increase in L/T debt 45 0 (30) 0 Increase in S/T debt (131) (184) (255) (385) Cash Dividend Paid 0 0 0 0 Dir& Emp Bonus Paid 0 0 0 0 Issuance of stock 0 0 0 0 Balance Sheet Other adjustments 6 0 0 0 NT$mn (Years End Dec ) 2014 2015E 2016E 2017E Exchange rate adjustment 780 940 1,052 1,187 Cash Net change in cash 158 160 112 135 0 0 0 0 Mkt Securities 174 163 231 276 AR/NR 46 47 67 81 Inventory Financial Ratios 6 6 6 6 Other 2014 2015E 2016E 2017E 1,006 1,156 1,356 1,550 Current Assets Growth(%) 0 0 0 0 Long-term investments Turnover 44.9 19.7 44.5 21.6 13 18 27 38 Fixed assets Operating profits 67.7 24.3 44.5 21.2 1 1 1 1 Deferred assets EPS 44.9 20.7 41.4 21.1 23 23 23 23 Other assets Margins (%) 1,043 1,199 1,407 1,613 Total Assets Gross Margin 55.0 55.4 56.3 56.4 45 45 15 15 S/T borrowings Operating Margin 35.7 37.1 37.1 37.0 34 35 50 60 AP/NP Pretax Margin 38.0 37.6 37.6 37.4 103 103 103 103 Other ST liabilities Net Margin 32.8 33.1 32.4 32.2 0 0 0 0 LT debt Return (%) 0 0 0 0 35.8 36.1 42.5 43.4 Other LT liabilities ROAE 263 263 263 263 30.1 30.2 36.8 38.4 Common shares ROAA 182 183 168 178 Total Liabilities Gearing (%) 305 305 305 305 (83.4) (88.1) (83.6) (81.7) Additional capital Net Debt/Equity 432 587 811 1,006 21.2 18.0 13.5 12.4 Retained earning Liabilities/Equity (139) (139) (139) (139) Other shareholders' equity Ratios (X) 861 1,015 1,239 1,435 5.5 6.3 8.1 8.7 Total Equity Current ratio 5.2 6.0 7.6 8.2 Total Liab. & Shrhldr's Equity 1,043 1,199 1,407 1,613 Quick ratio Others E = Morgan Stanley Research Estimates AR/NR Turnover (days) 59 58 57 56 Source: Morgan Stanley Research, Company Data Inventory Turnover (days) 38 38 38 38 AP Turnover (days) 28 28 28 28 Cash Conversion (days) 69 68 67 66 Source: Morgan Stanley Research estimates 46

Global Technology| July 22, 2015 IInnootteerraa EExxhhiibbiitt 2211:: Inotera: Quarterly and annual earnings estimates NT$ in million 1Q14 2Q14 3Q14 4Q14 1Q15 2Q15E 3Q15E 4Q15E 2013 2014 2015E 2016E 2017E Total Revenues 20,224 21,450 20,267 20,629 18,454 15,807 12,542 12,475 58,992 82,571 59,278 71,372 63,533 Sequential Change -2.2% 6.1% -5.5% 1.8% -10.5% -14.3% -20.7% -0.5% Change vs Year Ago 134.3% 68.3% 19.6% -0.2% -8.8% -26.3% -38.1% -39.5% 67.1% 40.0% -28.2% 20.4% -11.0% Cost of Sales (9,291) (9,429) (9,443) (9,277) (9,263) (10,281) (10,543) (11,217) (37,086) (37,440) (41,304) (47,871) (41,239) Percent of Revenues 45.9% 44.0% 46.6% 45.0% 50.2% 65.0% 84.1% 89.9% 62.9% 45.3% 69.7% 67.1% 64.9% Variable costs (5,681) (5,884) (5,927) (5,550) (5,747) (6,137) (6,296) (6,135) (22,051) (23,042) (24,315) (25,992) (25,792) as a % of revenue 28.1% 27.4% 29.2% 26.9% 31.1% 38.8% 50.2% 49.2% 37.4% 27.9% 41.0% 36.4% 40.6% Dep. & Amort. expense (3,610) (3,545) (3,516) (3,728) (3,516) (4,144) (4,247) (5,082) (15,035) (14,399) (16,989) (21,879) (15,447) as a % of COGS 38.9% 37.6% 37.2% 40.2% 38.0% 40.3% 40.3% 45.3% 40.5% 38.5% 41.1% 45.7% 37.5% as a % of revenue 17.9% 16.5% 17.3% 18.1% 19.1% 26.2% 33.9% 40.7% 25.5% 17.4% 28.7% 30.7% 24.3% Gross Margin 10,933 12,021 10,824 11,352 9,190 5,527 1,999 1,258 21,907 45,130 17,974 23,501 22,294 Percent of Revenues 54.1% 56.0% 53.4% 55.0% 49.8% 35.0% 15.9% 10.1% 37.1% 54.7% 30.3% 32.9% 35.1% Incremental Margins NM 88.7% 101.2% 145.7% 99.4% 138.4% 108.0% 1106.8% Total Opex (209) (306) (312) (519) (487) (280) (280) (280) (608) (1,346) (1,327) (1,300) (1,200) Percent of Revenues 1.0% 1.4% 1.5% 2.5% 2.6% 1.8% 2.2% 2.2% 1.0% 1.6% 2.2% 1.8% 1.9% R&D (138) (233) (235) (422) (403) (200) (200) (200) (299) (1,028) (1,003) (980) (880) Percent of Revenues 0.7% 1.1% 1.2% 2.0% 2.2% 1.3% 1.6% 1.6% 0.5% 1.2% 1.7% 1.4% 1.4% Sales and Marketing 0 0 0 0 0 0 0 0 0 0 0 0 0 Percent of Revenues 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% General and Admin (71) (73) (77) (97) (84) (80) (80) (80) (309) (318) (324) (320) (320) Percent of Revenues 0.4% 0.3% 0.4% 0.5% 0.5% 0.5% 0.6% 0.6% 0.5% 0.4% 0.5% 0.4% 0.5% Operating Income 10,724 11,715 10,513 10,833 8,703 5,247 1,719 978 21,299 43,784 16,647 22,201 21,094 Percent of Revenues 53.0% 54.6% 51.9% 52.5% 47.2% 33.2% 13.7% 7.8% 36.1% 53.0% 28.1% 31.1% 33.2% EBITDA 14,334 15,260 14,029 14,560 12,220 9,391 5,966 6,060 36,334 58,183 33,636 44,080 36,542 EBITDA margin 70.9% 71.1% 69.2% 70.6% 66.2% 59.4% 47.6% 48.6% 61.6% 70.5% 56.7% 61.8% 57.5% Non-operating Income(Loss) 522 (1,100) 1,116 2,362 (491) (57) (57) (65) 71 2,766 (669) (193) (391) Forex gain 723 (723) 941 2,363 0 0 0 0 443 3,304 0 0 0 Interest expense (275) (254) (231) (145) (109) (94) (94) (79) (1,091) (906) (376) (343) (541) Interest income 45 85 87 68 61 38 38 14 20 150 150 150 150 Net Investment Income 0 0 0 0 0 0 0 0 0 0 0 0 0 Other income (expense) 29 (208) 320 77 (443) 0 0 0 700 217 (443) 0 0 Profit Before Taxes 11,246 10,615 11,629 13,195 8,212 5,190 1,662 913 21,370 46,551 15,978 22,008 20,703 Percent of Revenues 55.6% 49.5% 57.4% 64.0% 44.5% 32.8% 13.3% 7.3% 36.2% 56.4% 27.0% 30.8% 32.6% Taxes 0 0 (0) 6,228 (943) (597) (191) (105) 0 6,228 (1,836) (2,735) (2,691) Tax Rate 0.0% 0.0% 0.0% -47.2% 11.5% 11.5% 11.5% 11.5% 0.0% -13.4% 11.5% 12.4% 13.0% Reported Income (TW GAAP) 11,246 10,615 11,629 19,423 7,269 4,593 1,471 808 21,370 52,779 14,142 19,272 18,012 Percent of Revenues 55.6% 49.5% 57.4% 94.2% 39.4% 29.1% 11.7% 6.5% 36.2% 63.9% 23.9% 27.0% 28.3% Change vs Year Ago NM NM NM NM NM -56.7% -87.3% -95.8% NM 147.0% -73.2% 36.3% -6.5% Reported EPS (NT$, TW GAAP) 1.85 1.71 1.85 3.05 1.11 0.70 0.22 0.12 3.76 8.47 2.16 2.95 2.75 Sequential Change -3.1% -7.3% 7.6% 65.5% -63.6% -36.8% -68.0% -45.1% NM 125.5% -74.5% 36.3% -6.5% Source: Morgan Stanley Research estimates 47

Global Technology| July 22, 2015 TTrriippoodd EEaarrnniinnggss RReevviissiioonn && VVaalluuaattiioonn MMeetthhooddoollooggyy We lower our 2015/2016/2017 EPS estimates by 7.2%, 9.2%, and 6.5%, to NT$4.60, NT$5.03, and NT$5.37, respectively, on the back of server and PC related end demand weakness. Our price target is thus reduced to NT$52.5, based on our residual income valuation model. This implies 2015e P/E of 11.4x. We continue to assume a 9% cost of equity, 5% intermediate-term growth rate, and 3% terminal growth rate. EExxhhiibbiitt 2222:: Tripod: Earnings Estimate Revision Source: Morgan Stanley Research estimates Scenario Analysis Revision We lower our bull case value to NT$72.8 from NT$78.6 to reflect our earnings revision from server demand weakness. This implies a 15.8x 2015e P/E. We also lower our bear case value to NT$40.0, from NT$41.2, implying 8.7x 2015e P/E. EExxhhiibbiitt 2233:: Tripod: Residual Income Valuation , 2015-25E Source: Company data, Morgan Stanley Research estimates 48

Global Technology| July 22, 2015 TTrriippoodd RRiisskk RReewwaarrdd Balanced Risk Reward WWhhyy EEqquuaall--wweeiigghhtt?? We now expect Tripod's 2015 profits to decline slightly given further weakness for the PC segment in 2Q and growth deceleration in server demand. Its dedicated efforts to diversify end exposure to the automotive and server/networking segments remain ongoing. Concerns are mainly its exposure to PC/NB segments and slower market share gain. Valuation looks fair at 12.0x 2015 P/E vs peers' 9- 13x. KKeeyy VVaalluuee DDrriivveerrss Market share gains to expand revenue stream. Potential new customer additions. Source: Thomson Reuters, Morgan Stanley Research Successful product mix upgrade through more automotive, server, and smartphone/tablet PC Derived from base-case scenario. Price Target NT$52.5 projects. Greater-than-expected rebound and market share expansion Bull NT$72.8 PPootteennttiiaall CCaattaallyyssttss merits: Better-than-expected end-demand strength and 15.8x 2015e P/E continuous market share gain should lead to revenue growth of Positive sell-through data of TFT LCD TV and 12% YoY in 2015 and 13% YoY in 2016 and thus greater smartphones/tablet PCs. profitability - sustainable OPM at 8-9% over the two years. Earlier-than-expected PC segment recovery. Penetration into new customers with mass volume. Profit sustained amid business transition: We expect Tripod to Base NT$52.5 keep its revenue flat to up slightly in 2015-16e by diversifying 11.4x 2015e P/E RRiisskkss ttoo AAcchhiieevviinngg PPrriiccee TTaarrggeett end-product exposure to automotive and server applications. Operating margin should stay at 6% in 2015-16e. Upside risks Weaker demand; greater ASP pressure: Slower shipments amid Bear NT$40.0 Strong rebound from TFT/DRAM/NB/TV business subdued demand also lead to greater ASP pressure on 8.7x 2015e P/E competition, resulting in revenue YoY decline in 2015-16 with Better-than-expected yield improvement lower OPM at 4-5%. Less price erosion with eased competition Greater-than-expected customers share gain Favorable forex and raw material cost changes. Downside risks Worse-than-expected underlying end demand in TFT, DRAM, NB, and HDD Slower-than-expected yield improvement Severe price competition on HDI Market share loss due to competitive pricing environment Unfavorable materials and forex movements. 49

Global Technology| July 22, 2015 TTrriippoodd FFiinnaanncciiaall SSuummmmaarryy EExxhhiibbiitt 2244:: Tripod: Quarterly Earnings Summary, 2014-16E Source: Company data, Morgan Stanley Research; E = Morgan Stanley Research estimates 50

Global Technology| July 22, 2015 EExxhhiibbiitt 2255:: Tripod: Consolidated Financial Summary Source: Company data, Morgan Stanley Research; E = Morgan Stanley Research estimates 51

Global Technology| July 22, 2015 52

Global Technology| July 22, 2015 Disclosure Section The information and opinions in Morgan Stanley Research were prepared by Morgan Stanley & Co. LLC, and/or Morgan Stanley C.T.V.M. S.A., and/or Morgan Stanley Mexico, Casa de Bolsa, S.A. de C.V., and/or Morgan Stanley Canada Limited. As used in this disclosure section, "Morgan Stanley" includes Morgan Stanley & Co. LLC, Morgan Stanley C.T.V.M. S.A., Morgan Stanley Mexico, Casa de Bolsa, S.A. de C.V., Morgan Stanley Canada Limited and their affiliates as necessary. For important disclosures, stock price charts and equity rating histories regarding companies that are the subject of this report, please see the Morgan Stanley Research Disclosure Website at www.morganstanley.com/researchdisclosures, or contact your investment representative or Morgan Stanley Research at 1585 Broadway, (Attention: Research Management), New York, NY, 10036 USA. For valuation methodology and risks associated with any price targets referenced in this research report, please contact the Client Support Team as follows: US/Canada +1 800 303-2495; Hong Kong +852 2848-5999; Latin America +1 718 754-5444 (U.S.); London +44 (0)20-7425-8169; Singapore +65 6834-6860; Sydney +61 (0)2-9770-1505; Tokyo +81 (0)3-6836-9000. Alternatively you may contact your investment representative or Morgan Stanley Research at 1585 Broadway, (Attention: Research Management), New York, NY 10036 USA. Analyst Certification The following analysts hereby certify that their views about the companies and their securities discussed in this report are accurately expressed and that they have not received and will not receive direct or indirect compensation in exchange for expressing specific recommendations or views in this report: Charlie Chan; Grace Chen; James E Faucette; Katy L. Huberty, CFA; Shawn Kim; Jasmine Lu; Joseph Moore; Brian Nowak, CFA; Shoji Sato; Sharon Shih; Keith Weiss, CFA. Unless otherwise stated, the individuals listed on the cover page of this report are research analysts. Global Research Conflict Management Policy Morgan Stanley Research has been published in accordance with our conflict management policy, which is available at www.morganstanley.com/institutional/research/conflictpolicies. Important US Regulatory Disclosures on Subject Companies The following analyst or strategist (or a household member) owns securities (or related derivatives) in a company that he or she covers or recommends in Morgan Stanley Research: Jasmine Lu - Wistron Corporation(common or preferred stock). As of June 30, 2015, Morgan Stanley beneficially owned 1% or more of a class of common equity securities of the following companies covered in Morgan Stanley Research: Amazon.com Inc, Apple, Inc., Broadcom Corporation, Facebook Inc, Infoblox Inc, Red Hat, Inc., Samsung Electronics, SanDisk Corporation, SK Hynix, Yandex NV. Within the last 12 months, Morgan Stanley managed or co-managed a public offering (or 144A offering) of securities of Amazon.com Inc, AT&T, Inc., Broadcom Corporation, Cisco Systems Inc, Juniper Networks Inc, Micron Technology Inc., Red Hat, Inc., Texas Instruments, Verizon Communications. Within the last 12 months, Morgan Stanley has received compensation for investment banking services from Amazon.com Inc, Arista Networks, AT&T, Inc., Baidu Inc, Broadcom Corporation, Cisco Systems Inc, Facebook Inc, Google, Juniper Networks Inc, Micron Technology Inc., Red Hat, Inc., Texas Instruments, Verizon Communications. In the next 3 months, Morgan Stanley expects to receive or intends to seek compensation for investment banking services from A10 Networks, Inc., Altera Corporation, Amazon.com Inc, Apple, Inc., Arista Networks, AT&T, Inc., Avago Technologies Ltd, Baidu Inc, Broadcom Corporation, Cavium Inc, Cisco Systems Inc, Citrix Systems Inc, EMC Corp., F5 Networks Inc, Facebook Inc, Google, Hewlett-Packard, IBM, Infoblox Inc, Inotera Memories, Inc., Inphi Corporation, Intel Corporation, Juniper Networks Inc, Marvell Technology Group Ltd, Micron Technology Inc., Microsoft, NetApp Inc, Nimble Storage, NVIDIA Corp., QLogic Corporation, Red Hat, Inc., Samsung Electronics, SanDisk Corporation, Seagate Technology, SK Hynix, Tencent Holdings Ltd., Texas Instruments, Verizon Communications, VMware Inc, Western Digital, Xilinx, Yahoo! Inc, Yandex NV. Within the last 12 months, Morgan Stanley has received compensation for products and services other than investment banking services from Altera Corporation, Apple, Inc., AT&T, Inc., Avago Technologies Ltd, Baidu Inc, Cisco Systems Inc, Citrix Systems Inc, EMC Corp., Google, Hewlett-Packard, Hon Hai Precision, IBM, Intel Corporation, Juniper Networks Inc, Marvell Technology Group Ltd, Micron Technology Inc., Microsoft, NetApp Inc, NVIDIA Corp., Quanta Computer Inc., Red Hat, Inc., SanDisk Corporation, Seagate Technology, Tencent Holdings Ltd., Texas Instruments, Tripod Technology, Verizon Communications, Yandex NV. Within the last 12 months, Morgan Stanley has provided or is providing investment banking services to, or has an investment banking client relationship with, the following company: A10 Networks, Inc., Altera Corporation, Amazon.com Inc, Apple, Inc., Arista Networks, AT&T, Inc., Avago Technologies Ltd, Baidu Inc, Broadcom Corporation, Cavium Inc, Cisco Systems Inc, Citrix Systems Inc, EMC Corp., F5 Networks Inc, Facebook Inc, Google, Hewlett- Packard, IBM, Infoblox Inc, Inotera Memories, Inc., Inphi Corporation, Intel Corporation, Juniper Networks Inc, Marvell Technology Group Ltd, Micron Technology Inc., Microsoft, NetApp Inc, Nimble Storage, NVIDIA Corp., QLogic Corporation, Red Hat, Inc., Samsung Electronics, SanDisk Corporation, Seagate Technology, SK Hynix, Tencent Holdings Ltd., Texas Instruments, Verizon Communications, VMware Inc, Western Digital, Xilinx, Yahoo! Inc, Yandex NV. Within the last 12 months, Morgan Stanley has either provided or is providing non-investment banking, securities-related services to and/or in the past has entered into an agreement to provide services or has a client relationship with the following company: Altera Corporation, Amazon.com Inc, Apple, Inc., AT&T, Inc., Avago Technologies Ltd, Baidu Inc, Broadcom Corporation, Cisco Systems Inc, Citrix Systems Inc, EMC Corp., F5 Networks Inc, Google, Hewlett-Packard, Hon Hai Precision, IBM, Intel Corporation, Juniper Networks Inc, Marvell Technology Group Ltd, Micron Technology Inc., Microsoft, NetApp Inc, NVIDIA Corp., QLogic Corporation, Quanta Computer Inc., Red Hat, Inc., SanDisk Corporation, Seagate Technology, Tencent Holdings Ltd., Texas Instruments, Tripod Technology, Verizon Communications, Wistron Corporation, Xilinx, Yahoo! Inc, Yandex NV. An employee, director or consultant of Morgan Stanley is a director of AT&T, Inc., EMC Corp., Facebook Inc, Hewlett-Packard, IBM, Verizon Communications. This person is not a research analyst or a member of a research analyst's household. Morgan Stanley & Co. LLC makes a market in the securities of A10 Networks, Inc., Altera Corporation, Amazon.com Inc, Apple, Inc., Arista Networks, AT&T, Inc., Avago Technologies Ltd, Baidu Inc, Broadcom Corporation, Cavium Inc, Cisco Systems Inc, Citrix Systems Inc, EMC Corp., F5 Networks Inc, Facebook Inc, Google, Hewlett-Packard, IBM, Infoblox Inc, Inphi Corporation, Intel Corporation, Juniper Networks Inc, Marvell Technology Group Ltd, Micron Technology Inc., Microsoft, NetApp Inc, NVIDIA Corp., QLogic Corporation, Red Hat, Inc., SanDisk Corporation, Seagate Technology, Texas Instruments, Verizon Communications, VMware Inc, Western Digital, Xilinx, Yahoo! Inc, Yandex NV. The equity research analysts or strategists principally responsible for the preparation of Morgan Stanley Research have received compensation based upon various factors, including quality of research, investor client feedback, stock picking, competitive factors, firm revenues and overall investment banking revenues. Morgan Stanley and its affiliates do business that relates to companies/instruments covered in Morgan Stanley Research, including market making, providing liquidity and specialized trading, risk arbitrage and other proprietary trading, fund management, commercial banking, extension of credit, investment services and investment banking. Morgan Stanley sells to and buys from customers the securities/instruments of companies covered in Morgan Stanley Research on a principal basis. Morgan Stanley may have a position in the debt of the Company or instruments discussed in this report. Certain disclosures listed above are also for compliance with applicable regulations in non-US jurisdictions. STOCK RATINGS Morgan Stanley uses a relative rating system using terms such as Overweight, Equal-weight, Not-Rated or Underweight (see definitions below). Morgan Stanley does not assign ratings of Buy, Hold or Sell to the stocks we cover. Overweight, Equal-weight, Not-Rated and Underweight are not the equivalent of buy, hold and sell. Investors should carefully read the definitions of all ratings used in Morgan Stanley Research. In addition, since Morgan Stanley 53

Global Technology| July 22, 2015 Research contains more complete information concerning the analyst's views, investors should carefully read Morgan Stanley Research, in its entirety, and not infer the contents from the rating alone. In any case, ratings (or research) should not be used or relied upon as investment advice. An investor's decision to buy or sell a stock should depend on individual circumstances (such as the investor's existing holdings) and other considerations. Global Stock Ratings Distribution (as of June 30, 2015) For disclosure purposes only (in accordance with NASD and NYSE requirements), we include the category headings of Buy, Hold, and Sell alongside our ratings of Overweight, Equal-weight, Not-Rated and Underweight. Morgan Stanley does not assign ratings of Buy, Hold or Sell to the stocks we cover. Overweight, Equal-weight, Not-Rated and Underweight are not the equivalent of buy, hold, and sell but represent recommended relative weightings (see definitions below). To satisfy regulatory requirements, we correspond Overweight, our most positive stock rating, with a buy recommendation; we correspond Equal-weight and Not-Rated to hold and Underweight to sell recommendations, respectively. COVERAGE UNIVERSE INVESTMENT BANKING CLIENTS (IBC) STOCK RATING CATEGORY COUNT % OF TOTAL COUNT % OF TOTAL % OF RATING IBC CATEGORY Overweight/Buy 1183 35% 315 43% 27% Equal-weight/Hold 1456 44% 336 45% 23% Not-Rated/Hold 93 3% 9 1% 10% Underweight/Sell 613 18% 79 11% 13% TOTAL 3,345 739 Data include common stock and ADRs currently assigned ratings. Investment Banking Clients are companies from whom Morgan Stanley received investment banking compensation in the last 12 months. Analyst Stock Ratings Overweight (O). The stock's total return is expected to exceed the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months. Equal-weight (E). The stock's total return is expected to be in line with the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months. Not-Rated (NR). Currently the analyst does not have adequate conviction about the stock's total return relative to the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months. Underweight (U). The stock's total return is expected to be below the average total return of the analyst's industry (or industry team's) coverage universe, on a risk-adjusted basis, over the next 12-18 months. Unless otherwise specified, the time frame for price targets included in Morgan Stanley Research is 12 to 18 months. Analyst Industry Views Attractive (A): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be attractive vs. the relevant broad market benchmark, as indicated below. In-Line (I): The analyst expects the performance of his or her industry coverage universe over the next 12-18 months to be in line with the relevant broad market benchmark, as indicated below. Cautious (C): The analyst views the performance of his or her industry coverage universe over the next 12-18 months with caution vs. the relevant broad market benchmark, as indicated below. Benchmarks for each region are as follows: North America - S&P 500; Latin America - relevant MSCI country index or MSCI Latin America Index; Europe - MSCI Europe; Japan - TOPIX; Asia - relevant MSCI country index or MSCI sub-regional index or MSCI AC Asia Pacific ex Japan Index. Stock Price, Price Target and Rating History (See Rating Definitions) 54

Global Technology| July 22, 2015 Important Disclosures for Morgan Stanley Smith Barney LLC Customers Important disclosures regarding the relationship between the companies that are the subject of Morgan Stanley Research and Morgan Stanley Smith 55

Global Technology| July 22, 2015 Barney LLC or Morgan Stanley or any of their affiliates, are available on the Morgan Stanley Wealth Management disclosure website at www.morganstanley.com/online/researchdisclosures. For Morgan Stanley specific disclosures, you may refer to www.morganstanley.com/researchdisclosures. Each Morgan Stanley Equity Research report is reviewed and approved on behalf of Morgan Stanley Smith Barney LLC. This review and approval is conducted by the same person who reviews the Equity Research report on behalf of Morgan Stanley. This could create a conflict of interest. Other Important Disclosures Morgan Stanley & Co. International PLC and its affiliates have a significant financial interest in the debt securities of Altera Corporation, Amazon.com Inc, Apple, Inc., AT&T, Inc., Baidu Inc, Broadcom Corporation, Cisco Systems Inc, EMC Corp., Facebook Inc, Google, Hewlett-Packard, IBM, Intel Corporation, Juniper Networks Inc, Micron Technology Inc., Microsoft, NetApp Inc, Samsung Electronics, SanDisk Corporation, SK Hynix, Tencent Holdings Ltd., Texas Instruments, Verizon Communications, Xilinx, Yahoo! Inc. Morgan Stanley is not acting as a municipal advisor and the opinions or views contained herein are not intended to be, and do not constitute, advice within the meaning of Section 975 of the Dodd-Frank Wall Street Reform and Consumer Protection Act. Morgan Stanley produces an equity research product called a "Tactical Idea." Views contained in a "Tactical Idea" on a particular stock may be contrary to the recommendations or views expressed in research on the same stock. This may be the result of differing time horizons, methodologies, market events, or other factors. For all research available on a particular stock, please contact your sales representative or go to Matrix at http://www.morganstanley.com/matrix. Morgan Stanley Research is provided to our clients through our proprietary research portal on Matrix and also distributed electronically by Morgan Stanley to clients. Certain, but not all, Morgan Stanley Research products are also made available to clients through third-party vendors or redistributed to clients through alternate electronic means as a convenience. For access to all available Morgan Stanley Research, please contact your sales representative or go to Matrix at http://www.morganstanley.com/matrix. Any access and/or use of Morgan Stanley Research is subject to Morgan Stanley's Terms of Use (http://www.morganstanley.com/terms.html). By accessing and/or using Morgan Stanley Research, you are indicating that you have read and agree to be bound by our Terms of Use (http://www.morganstanley.com/terms.html). In addition you consent to Morgan Stanley processing your personal data and using cookies in accordance with our Privacy Policy and our Global Cookies Policy (http://www.morganstanley.com/privacy_pledge.html), including for the purposes of setting your preferences and to collect readership data so that we can deliver better and more personalized service and products to you. To find out more information about how Morgan Stanley processes personal data, how we use cookies and how to reject cookies see our Privacy Policy and our Global Cookies Policy (http://www.morganstanley.com/privacy_pledge.html). If you do not agree to our Terms of Use and/or if you do not wish to provide your consent to Morgan Stanley processing your personal data or using cookies please do not access our research. Morgan Stanley Research does not provide individually tailored investment advice. Morgan Stanley Research has been prepared without regard to the circumstances and objectives of those who receive it. Morgan Stanley recommends that investors independently evaluate particular investments and strategies, and encourages investors to seek the advice of a financial adviser. The appropriateness of an investment or strategy will depend on an investor's circumstances and objectives. The securities, instruments, or strategies discussed in Morgan Stanley Research may not be suitable for all investors, and certain investors may not be eligible to purchase or participate in some or all of them. Morgan Stanley Research is not an offer to buy or sell or the solicitation of an offer to buy or sell any security/instrument or to participate in any particular trading strategy. The value of and income from your investments may vary because of changes in interest rates, foreign exchange rates, default rates, prepayment rates, securities/instruments prices, market indexes, operational or financial conditions of companies or other factors. There may be time limitations on the exercise of options or other rights in securities/instruments transactions. Past performance is not necessarily a guide to future performance. Estimates of future performance are based on assumptions that may not be realized. If provided, and unless otherwise stated, the closing price on the cover page is that of the primary exchange for the subject company's securities/instruments. The fixed income research analysts, strategists or economists principally responsible for the preparation of Morgan Stanley Research have received compensation based upon various factors, including quality, accuracy and value of research, firm profitability or revenues (which include fixed income trading and capital markets profitability or revenues), client feedback and competitive factors. Fixed Income Research analysts', strategists' or economists' compensation is not linked to investment banking or capital markets transactions performed by Morgan Stanley or the profitability or revenues of particular trading desks. The "Important US Regulatory Disclosures on Subject Companies" section in Morgan Stanley Research lists all companies mentioned where Morgan Stanley owns 1% or more of a class of common equity securities of the companies. For all other companies mentioned in Morgan Stanley Research, Morgan Stanley may have an investment of less than 1% in securities/instruments or derivatives of securities/instruments of companies and may trade them in ways different from those discussed in Morgan Stanley Research. Employees of Morgan Stanley not involved in the preparation of Morgan Stanley Research may have investments in securities/instruments or derivatives of securities/instruments of companies mentioned and may trade them in ways different from those discussed in Morgan Stanley Research. Derivatives may be issued by Morgan Stanley or associated persons. With the exception of information regarding Morgan Stanley, Morgan Stanley Research is based on public information. Morgan Stanley makes every effort to use reliable, comprehensive information, but we make no representation that it is accurate or complete. We have no obligation to tell you when opinions or information in Morgan Stanley Research change apart from when we intend to discontinue equity research coverage of a subject company. Facts and views presented in Morgan Stanley Research have not been reviewed by, and may not reflect information known to, professionals in other Morgan Stanley business areas, including investment banking personnel. Morgan Stanley Research personnel may participate in company events such as site visits and are generally prohibited from accepting payment by the company of associated expenses unless pre-approved by authorized members of Research management. Morgan Stanley may make investment decisions or take proprietary positions that are inconsistent with the recommendations or views in this report. To our readers in Taiwan: Information on securities/instruments that trade in Taiwan is distributed by Morgan Stanley Taiwan Limited ("MSTL"). Such information is for your reference only. The reader should independently evaluate the investment risks and is solely responsible for their investment decisions. Morgan Stanley Research may not be distributed to the public media or quoted or used by the public media without the express written consent of Morgan Stanley. Information on securities/instruments that do not trade in Taiwan is for informational purposes only and is not to be construed as a recommendation or a solicitation to trade in such securities/instruments. MSTL may not execute transactions for clients in these securities/instruments. To our readers in Hong Kong: Information is distributed in Hong Kong by and on behalf of, and is attributable to, Morgan Stanley Asia Limited as part of its regulated activities in Hong Kong. If you have any queries concerning Morgan Stanley Research, please contact our Hong Kong sales representatives. Morgan Stanley is not incorporated under PRC law and the research in relation to this report is conducted outside the PRC. Morgan Stanley Research does not constitute an offer to sell or the solicitation of an offer to buy any securities in the PRC. PRC investors shall have the relevant qualifications to invest in such securities and shall be responsible for obtaining all relevant approvals, licenses, verifications and/or registrations from the relevant governmental authorities themselves. Morgan Stanley Research is disseminated in Brazil by Morgan Stanley C.T.V.M. S.A.; in Japan by Morgan Stanley MUFG Securities Co., Ltd. and, for Commodities related research reports only, Morgan Stanley Capital Group Japan Co., Ltd; in Hong Kong by Morgan Stanley Asia Limited (which accepts responsibility for its contents) and by Bank Morgan Stanley AG, Hong Kong Branch; in Singapore by Morgan Stanley Asia (Singapore) Pte. (Registration number 199206298Z) and/or Morgan Stanley Asia (Singapore) Securities Pte Ltd (Registration number 200008434H), regulated by the Monetary Authority of Singapore (which accepts legal responsibility for its contents and should be contacted with respect to any matters arising from, or in connection with, Morgan Stanley Research) and by Bank Morgan Stanley AG, Singapore Branch (Registration number T11FC0207F); in Australia to "wholesale clients" within the meaning of the Australian Corporations Act by Morgan Stanley Australia Limited A.B.N. 67 003 734 576, holder of Australian financial services license No. 233742, which accepts responsibility for its contents; in Australia to "wholesale clients" and "retail clients" within the meaning of the Australian Corporations Act by Morgan Stanley Wealth Management Australia Pty Ltd (A.B.N. 19 009 145 555, holder of Australian financial services license No. 56

Global Technology| July 22, 2015 240813, which accepts responsibility for its contents; in Korea by Morgan Stanley & Co International plc, Seoul Branch; in India by Morgan Stanley India Company Private Limited; in Indonesia by PT Morgan Stanley Asia Indonesia; in Canada by Morgan Stanley Canada Limited, which has approved of and takes responsibility for its contents in Canada; in Germany by Morgan Stanley Bank AG, Frankfurt am Main and Morgan Stanley Private Wealth Management Limited, Niederlassung Deutschland, regulated by Bundesanstalt fuer Finanzdienstleistungsaufsicht (BaFin); in Spain by Morgan Stanley, S.V., S.A., a Morgan Stanley group company, which is supervised by the Spanish Securities Markets Commission (CNMV) and states that Morgan Stanley Research has been written and distributed in accordance with the rules of conduct applicable to financial research as established under Spanish regulations; in the US by Morgan Stanley & Co. LLC, which accepts responsibility for its contents. Morgan Stanley & Co. International plc, authorized by the Prudential Regulatory Authority and regulated by the Financial Conduct Authority and the Prudential Regulatory Authority, disseminates in the UK research that it has prepared, and approves solely for the purposes of section 21 of the Financial Services and Markets Act 2000, research which has been prepared by any of its affiliates. Morgan Stanley Private Wealth Management Limited, authorized and regulated by the Financial Conduct Authority, also disseminates Morgan Stanley Research in the UK. Private UK investors should obtain the advice of their Morgan Stanley & Co. International plc or Morgan Stanley Private Wealth Management representative about the investments concerned. RMB Morgan Stanley (Proprietary) Limited is a member of the JSE Limited and regulated by the Financial Services Board in South Africa. RMB Morgan Stanley (Proprietary) Limited is a joint venture owned equally by Morgan Stanley International Holdings Inc. and RMB Investment Advisory (Proprietary) Limited, which is wholly owned by FirstRand Limited. Morgan Stanley Hong Kong Securities Limited is the liquidity provider/market maker for securities of Tencent Holdings Ltd. listed on the Stock Exchange of Hong Kong Limited. An updated list can be found on HKEx website: http://www.hkex.com.hk. The information in Morgan Stanley Research is being communicated by Morgan Stanley & Co. International plc (DIFC Branch), regulated by the Dubai Financial Services Authority (the DFSA), and is directed at Professional Clients only, as defined by the DFSA. The financial products or financial services to which this research relates will only be made available to a customer who we are satisfied meets the regulatory criteria to be a Professional Client. The information in Morgan Stanley Research is being communicated by Morgan Stanley & Co. International plc (QFC Branch), regulated by the Qatar Financial Centre Regulatory Authority (the QFCRA), and is directed at business customers and market counterparties only and is not intended for Retail Customers as defined by the QFCRA. As required by the Capital Markets Board of Turkey, investment information, comments and recommendations stated here, are not within the scope of investment advisory activity. Investment advisory service is provided exclusively to persons based on their risk and income preferences by the authorized firms. Comments and recommendations stated here are general in nature. These opinions may not fit to your financial status, risk and return preferences. For this reason, to make an investment decision by relying solely to this information stated here may not bring about outcomes that fit your expectations. The following companies do business in countries which are generally subject to comprehensive sanctions programs administered or enforced by the U.S. Department of the Treasury's Office of Foreign Assets Control ("OFAC") and by other countries and multi-national bodies: Samsung Electronics. The trademarks and service marks contained in Morgan Stanley Research are the property of their respective owners. Third-party data providers make no warranties or representations relating to the accuracy, completeness, or timeliness of the data they provide and shall not have liability for any damages relating to such data. The Global Industry Classification Standard (GICS) was developed by and is the exclusive property of MSCI and S&P. Morgan Stanley Research or portions of it may not be reprinted, sold or redistributed without the written consent of Morgan Stanley. Morgan Stanley Research, or any portion thereof may not be reprinted, sold or redistributed without the written consent of Morgan Stanley. © 2015 Morgan Stanley 57