EQUITY RESEARCH REPORT | 12/20/19

Recommendation: BUY CORPORATION (Ticker: NVDA) Business Rating: 6 Security Rating: 5

BUSINESS RATING SECURITY RATING

SELL BUY NEGATIVE POSITIVE d

POTENTIAL FOR RETURN RISK

LIMITED SIGNIFICANT LOW HIGH D Industry: Semiconductors Sector: Technology Price: $235.46 (12/19/19) Jarvis Rank: 42 (12/14/19)

(Data as of 12/19/19 unless specified)

Fwd (TTM) $12.7B Enterprise Value: $136.9B Market Cap: $144.1B Revenue: ($10.0B) 10.8x TTM Operating Revenue Growth Fwd (TTM) EV/Sales: 21.46% -9.76% (13.8x) Margin: (YoY): 28.4x Fwd (TTM) EV/EBITDA TTM FCF Margin: 36.4% RSI: 76.00 (55.4x) Share 52 week high $236.00 52 week low $124.46 +72.9% Performance YTD: Insider Transactions: Insiders have sold a net of 462.3k shares in the last twelve months

 WHY WE RATE NVIDIA (NVDA) A BUY • Nvidia is positioned strongly to capture exponential Grab-and-Go transformative secular trends in artificial intelligence (AI), cloud, THESIS internet of things (IoT) and autonomous vehicles An investment in Nvidia is a play • The company’s platform strategy involving CUDA and on the rapidly expanding demand developers is turning into the industry standard for AI/ML/DL for parallel processing computing and will lead to strong network effects that widen their moat power driven by gaming, artificial • Nvidia’s GPU lineup is moving towards 7nm architectures that intelligence, cloud transformation will likely result in a sharp surge in performance and autonomous vehicles. Nvidia is a market leading GPU and SOC • Strong profitability and healthy FCF margins; the company is a developer that is positioned well to cash cow, allowing for significant reinvestment in R&D, and the capture future secular trends in capture of new market opportunities industry demand. Technology • Gaming revenues are experiencing a down-cycle in demand; this advances, backed by successful doesn’t affect excellent long-term prospects, but it may lead to execution of their platform-centric strategies, are key to future growth. share price volatility • Jensen Huang is an excellent Founder/CEO and has led the company’s innovative transformations since its inception; we have high confidence in the current management • At Fwd EV/EBITDA of 28.4x, we believe the valuation presents a further upside to the share price and rate NVDA a BUY

“Seeking that can 1 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

 SUMMARY OF THE BUSINESS AND THE INDUSTRY ◼ Business 1 Nvidia is a leading semiconductor technology  company that engages in the design, development, Grab-and-Go and distribution of graphics processing units (GPUs*) TALKING POINTS for gaming and professional markets, systems on chips (SOC**) for the mobile computing and WHAT WE LIKE WHAT WE DON’T LIKE automotive market, and related software. The company has a platform strategy that brings together Proven pedigree in tech hardware, system software, programmable Highly volatile share capabilities and market- algorithms, libraries, systems and services to create price action leading position unique value for the markets it serves. The use cases for its GPUs include 3D graphics, virtual reality (VR), Platform-driven strategy Cyclicality and high-performance computing (HPC), artificial and network effects slowdown in gaming intelligence (AI) and machine learning (ML). through CUDA revenue The requirements of their various markets are diverse but are addressed by Nvidia with a single underlying Impressive cash flow Tech giants attempting their own generation and strong architecture, leveraging its GPUs and Compute chips; AMD’s data Unified Device Architecture (CUDA) as fundamental margins center wins building blocks. The architecture is programmable, with the company offering comprehensive software stacks developed either internally or by third-party developers and partners.

*GPUs are processors (chips) designed to rapidly manipulate and alter memory to accelerate the creation of images. While originally designed for graphics, they excel at processes that involve specialized operations performed at high speed, which differs from Central Processing Units (CPUs) that are better suited for serial processing that handles multitasking well. 2,3 **SOC are chips that integrate all the components of a computer or other electronic system on a single substrate. These components could include CPUs, memory, ports and storage.

1 https://investor.Nvidia.com/home/default.aspx 2 https://www.pcmag.com/encyclopedia/term/43886/gpu 3 https://blogs.Nvidia.com/blog/2009/12/16/whats-the-difference-between-a-cpu-and-a-gpu/

“Seeking Stocks that can 2 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

The company has two reportable segments – GPU and Processor; the chips are based on the same underlying architecture but encompass different brands and products catered to targeted use cases. GPU • GeForce: GPUs for PC gaming and mainstream PCs o GPUs are offered in a variety of price ranges and models depending on capabilities • GeForce NOW: Cloud-based, game-streaming service • : GPUs for design professionals working in computer-aided design, video editing, special effects and other creative applications • Tesla: GPUs for AI utilizing deep learning (DL) and accelerated computing by leveraging parallel computing capabilities • GRID: Cloud and datacenter solutions with server-side graphics, managing and monitoring capabilities for Virtual PCs and Virtual Apps • DGX: Fully integrated workstation solutions for AI development with GPU-optimized software and enterprise-grade support

Tegra • Tegra is Nvidia’s series of SOC solutions for smartphones, mobile devices, digital assistants and more; Tegra has recently evolved to offer comprehensive SOCs for DRIVE and SHIELD • DRIVE AGX: Computer platform that offers autonomous car and driver assistance functionality provided by DL o These include the DRIVE AGX XavierTM and PegasusTM systems that cater to levels 2 and 3 and levels 4 and 5 * of automation, respectively o Partners for the AGX program include manufacturers such as Audi, Mercedes-Benz, Toyota, Volvo and Volkswagen • Clara AGX: Healthcare application framework for AI-powered imaging and genomics o Includes full-stack, GPU-accelerated libraries, Software Development Kits (SDKs), and reference applications for developers, researchers and scientists • SHIELD: Devices and services catering to home entertainment, AI and gaming • Jetson AGX: Series of embedded computing boards utilizing Tegra technology designed for accelerated ML applications and sold to developers

*Accepted industry standards for advancement in autonomous driving capabilities: Level 0 – No Automation; Level 1 – Driver Assistance; Level 2 – Partial Automation; Level 3 – Conditional Automation; Level 4 – High Automation; Level 5 – Full Automation

Business Strategy Nvidia cites the following key strategies that shape its overall business approach: • Advancing the GPU computing platform • Extending technology and platform leadership in AI • Extending technology and platform leadership in Visual Computing • Advancing the Autonomous Vehicle Platform • Leveraging Intellectual Property

“Seeking Stocks that can 3 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

LB•LOGIC The GPU business is highly R&D oriented, and performance vastly improves every year or two, keeping up with Moore’s Law*. Nvidia’s attempt at anchoring their hardware with a platform model, development community and partnerships is a strong approach that we believe expands their moat through network effects.*

Sales and Marketing Nvidia’s sales strategy involves working with end customers and industry ecosystems through its partner network. Sales and marketing teams are located across global markets and work closely with original equipment manufacturers (OEMs), original device manufacturers (ODMs), system builders, retailers/distributors, internet and cloud service providers, automotive manufacturers and suppliers, start- ups, and other participants. To promote the development of applications optimized for their GPUs and CUDA, the company seeks and establishes relationships with the developer community. Their initiatives encourage the development of AI frameworks, SDKs, APIs and other applications for their platform. Revenue growth across key segments

(Source: 2019 Nvidia Day Presentation) Gaming contributes the most to company revenues, while Datacenter applications are Nvidia’s fastest- growing market segment, driven by cloud infrastructure transformation and AI use cases.

*Moore’s Law refers to Moore's perception that the number of transistors on a microchip doubles every two years, though the cost of computers is halved. Moore's Law states that we can expect the speed and capability of our computers to increase every few years, and we will pay less for them.

“Seeking Stocks that can 4 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

◼ Outlook/Estimates in $ millions Q2 FY20E Q2 FY20A Q3 FY20E Q3 FY20A Q4 FY20E Revenue 2,550 2,579 2,900 3,014 2,950 Non-GAAP Gross Margin 59.2% 60.1% 62.0% 64.1% 64.5% Non-GAAP Operating Expenses 765 749 765 774 805

◼ Industry/Competition Industry trends4,5 The semiconductor industry experienced explosive growth during the past bull market, driven by strong secular and cyclical trends. With an increasing number of devices, the growth of gaming, high-performance computing, imaging and visualization software, cloud and data infrastructure, ML use cases, IoT, blockchain, and autonomous driving, the industry has benefitted from trends in innnovation since the first advent of the microprocessor. GPUs were first invented and named by Nvidia for their specialized use cases in graphics; their parallel computing capabilities were only later applied to other use cases, and the company remains well-positioned for the rise of autonomous vehicles, AI, and ML, and other exponential growth trends.

Cyclical trends: The GPU industry experienced rapid revenue expansion from 2014-2018 as favorable economic conditions led to increased consumer spending and businesses upgrading their hardware. The cryptocurrency bubble, in particular, fueled the sale of GPUs as they made for excellent mining hardware due to their natural parallel computing features. This led to inflated prices in 2017. 2018 saw slowing economic growth, and spending on GPUs subsided with a downcycle in gaming GPU sales continuing into 2019.

4 https://www.jagcapm.com/where-are-we-in-the-semiconductor-cycle/ 5 https://www.forbes.com/sites/jimhandy/2014/05/28/the-3-reasons-semiconductor-experience-revenue-cycles/

“Seeking Stocks that can 5 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

Secular trends: Driven by demand for graphics related cloud computing, data center applications for GPUs has fueled the secular trend of GPU adoption in recent times, with autonomous vehicles, AI and ML, and cryptocurrencies adding to the adoption boost. With several opportunities for catalyzed growth and newer use case markets in their nascent stage, the industry is well-positioned for dynamic growth across a variety of markets.

LB•LOGIC We believe the expansion of the industry for chip manufacturers and GPUs by extension over the long term is inevitable as requirements for more computing power and applications continue to increase.

Nvidia’s Key Markets Nvidia’s key market segments include Gaming, Data Centers, Professional Visualization, and Automotive. • Gaming GPU Market The gaming segment is the largest contributor of revenue across the company’s current addressable markets. Nvidia has been the clear leader in GPUs for consumer gaming purposes for most of its history, including in recent years, and has therefore enjoyed significant pricing power. The gaming GPU market, like other cyclical industries, saw strong revenue growth until 2018. Cryptocurrency mining had a huge impact on Nvidia’s GPU sales, which saw temporarily accelerated growth rates during the cryptocurrency bubble in late 2017. Revenue growth slowed, as did GPU sales, leading to a sharp correction for the stock in late 2018.

LB•LOGIC Cryptocurrency enthusiasts bought Nvidia GPUs during the bubble as the appropriate hardware for mining. Since then, several alternatives have arisen, catered specifically for mining applications, many of which use crypto-specific Application Specific Integrated Circuits (ASICs) that are optimized for mining applications. We believe the 2017 surge was inorganic and is unlikely to repeat.

• Professional Visualization Market The market comprises GPUs for workstation applications that could include visual effects, image rendering, video production, animation and other creative uses that require parallel computing features. • Data Center GPU Market6 The data center market consists of heavy workload applications such as Scientific Computing, Data Analytics, and Artificial Intelligence (ML and DL) used by enterprises, research and higher education institutions, and cloud service providers such as AWS, Azure and GCP. The demand for computing power to cater to these applications has surged and so have Nvidia’s data center revenues.

6 https://searchdatacenter.techtarget.com/answer/What-does-GPU-hardware-do-in-the-data-center

“Seeking Stocks that can 6 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

• Automotive Market7 Auto manufacturers have invested heavily in self-driving technology and DL applications to increase the capabilities of vehicles. As of 2019, regulations allow the sale of Level 3 (conditional automation) vehicles to consumers. As Level 4 and 5 (high and full automation) become mainstream, and refulations catch up, the demand for SOCs to power the core technologies enabling autonomy will see a surge. Technologies such as Lidar, Radar and Ultrasound are currently used in combination with cameras and other equipment to process data in real-time for vehicles and train proprietary AI systems through ML and DL. Once proficient, those technologies will be deployed in vehicles for sale.

LB•LOGIC We strongly believe that there will be an advanced built on a microprocessor architecture in every vehicle in the future. The market for Nvidia’s SOCs is still in its infancy.

Competition The only other major pure-play GPU developer that Nvidia currently competes with is AMD ().

AMD8,9 AMD is a leading manufacturer of CPUs and GPUs that competes with market leaders Intel and Nvidia. In terms of competitive products, AMD’s GPU offerings were considered inferior for several years, resulting in the company selling its products at large discounts to match Nvidia’s price to performance ratios while incurring excessive amounts of debt. The company has displayed strong signs by clawing back some market share from Intel and Nvidia over the past two years based on their early adoption of 7nm chip designs*. *7nm (7 nanometers) refers to the distance between transistor nodes on a chip. The industry has been developing chips at lower nanometer values over the years as the technology for enabling their design and manufacture has been improving. Fewer nanometers implies closer transistor packing on a die, resulting in generally higher efficiency and better performance. However, several other factors matter as well.

7 https://www.digitaltrends.com/cars/the-current-state-of-autonomous-vehicles/ 8 https://www.amd.com/en 9 https://www.theverge.com/2019/6/10/18660213/amd-ryzen-9-3950x-16-core-gaming-cpu-processor-e3-2019

“Seeking Stocks that can 7 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

• AMD vs Nvidia in the Gaming GPU market10

Market capture for GPUs is dependent heavily on product releases and timing. With vast improvements between every generation of GPU, market share for industry players keeps fluctuating. In Q1 and Q2 2019, AMD’s market share rose due to its 7 product releases featuring a 7nm architecture. The Radeon brand directly competes with Nvidia’s GeForce line. Nvidia followed the release with their GeForce RTX Super releases, an upgrade from GeForce RTX o AMD often reduces prices of its GPUs to compete with Nvidia’s offerings from a price/performance perspective o Nvidia maintains a ~70.0% share of the gaming market for discrete desktop GPUs o While the Desktop GPU market is a duopoly, Nvidia holds most of the market for discrete laptop GPUs o The latest generation of GeForce RTX chips feature , a rendering technique for mapping light across images, which the company claims provides a superior gaming experience; AMD doesn’t have a similar technology offering o Both companies have partnered with game developers to bundle and market games with sales of their new GPUs o Benchmarking, or testing, reveals that comparably priced offerings are well-matched, while Nvidia maintains a performance lead in the high-end market o While Nvidia used to offer a superior product lineup and held pricing power, AMD has released respectable products that can be considered competitive in 2019

10 https://www.jonpeddie.com/press-releases/

“Seeking Stocks that can 8 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

• AMD vs Nvidia for Professional Visualization11,12 Nvidia’s Quadro chips and AMD’s chips target the professional visualization market. Graphics professionals are typically less sensitive toward price, and determining which product has the best performance depends heavily on the application or software used. o Quadros are typically priced higher than Radeon Pros o The flagship Quadro workstation chips (branded ) offer significantly higher performance from a single graphics card when compared to Radeon’s high-end offerings; Titan cards are also priced extremely high o Nvidia has strong partnerships with Adobe, the world’s premier creative software provider for photography, video editing and other graphics-related workloads. Both companies have worked extensively to optimize Nvidia’s architecture with Adobe’s software. As a result, Nvidia chips perform far better than AMD’s with Adobe applications, which are considered the industry standard o Radeon products are priced more for budget-conscious and price-sensitive users

• AMD vs Nvidia in the Datacenter market13 AMD has been able to make a considerable impact in the data center market in recent quarters. Unlike Nvidia and Intel, AMD has a pedigree in both CPU and GPU design. Data centers that require an optimized combination of CPUs and GPUs can choose AMD as a single vendor. While AMD CPUs have surpassed Intel’s comparable offerings due to 7nm architectures, their data center CPU market has seen traction, which has led to upsell in their data center GPU business from network effects. o AMD’s latest architecture release for data center applications came in early 2019; Nvidia’s most recent architecture for such applications was released in 2017, which means that its due for an upgrade o AMD won large contracts during the year, such as powering ’s Stadia service (high- performance gaming over the cloud) o Nvidia features CUDA, which a significant number of developers and scientists using AI and ML rely upon; AMD has no such platform or software advantages, missing out on a sizeable fraction of the data center market o Stadia has recently been criticized for its slower service, raising questions about AMD’s ability to execute GPU installations

11 https://techgage.com/article/radeon-pro-vs-quadro-a-fresh-look-at-workstation-gpu-performance/8/ 12 https://www.pugetsystems.com/labs/articles/Premiere-Pro-CC-2019-AMD-Radeon-VII-vs-NVIDIA-GeForce-RTX-1395/ 13 https://www.networkworld.com/article/3438098/can-amd-convert-its-growing-gpu-presence-into-a-data-center-play.html

“Seeking Stocks that can 9 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

o Certain sources14,15 have reported that Nvidia is preparing to release 7nm GPUs in early 2020; AMD’s 7nm-architecture GPUs raised the company’s offering to par with with Nvidia; once the playing field is leveled, Nvidia might dominate once more

LB•LOGIC We believe AMD’s data center progress in 2019 can be attributed to the recency of their GPU releases in 2019. Nvidia’s Tesla T4s, their last standard for data center GPUs, was released in late 2018 and is due for an upgrade. Given the company’s pedigree in meaningful generational performance leaps, we expect strong releases soon featuring 7nm architectures.

Other Competition • Google’s Tensor Processing Units (TPUs)16 Google has developed a proprietary ASIC for DL and neural network applications optimized for the Google Cloud Platform (GCP). It claims to be significantly more capable than Nvidia’s Tesla V100 offerings at a lower cost. However, we found no indicatin they are selling their technologies to other data centers. This, however, still eats into Nvidia’s total addressable data center market as Google could use its own chips instead of buying Nvidia’s. GCP is one of the four major cloud infrastructure providers alongside AWS, Azure and Alibaba. • Other Big Tech attempts17,18 has developed its own CPUs (dubbed Graviton) but currently uses Nvidia’s V100s across their GPU cloud infrastructure within AWS; other news reports suggest that is building its own chips for its datacenters. • Tesla’s proprietary auto-chip19 Tesla has been at the forefront of autonomous driving and has developed an SoC/computer (alt. FSD or Full Self Driving computer) in-house that they claim is far ahead of the competition. They intend to use it for their own vehicles and calim it features Level 5 (full autonomy) capabilities.

LB•LOGIC Tesla has consistently paved the way in auto over the past few years and has succeeded where other manufacturers haven’t. Nvidia’s own releases of L4/L5 chips and partnerships with industry leaders such as Toyota and Daimler gives us confidence that their technology is robust and has a significatn market opportunity ahead.

14 https://www.notebookcheck.net/Rumor-Nvidia-Ampere-7nm-GPUs-to-launch-1H-2020.436973.0.html 15 https://wccftech.com/nvidia-7nm-ampere-gpus-landing-within-9-months-analyst-reports/ 16 https://cloud.google.com/tpu/ 17 https://www.forbes.com/sites/samshead/2019/02/19/facebook-plans-to-develop-its-own-ai-chips/#32e506a71837 18 https://www.wired.com/story/new-amazon-chips-cloud-computing/ 19 https://ark-invest.com/research/tesla-fsd

“Seeking Stocks that can 10 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

Competing Technologies and Emerging Trends • ASICs (Application Specific Integrated Circuits) Like Google’s TPUs, the industry is moving toward optimizing microprocessors for targeted functions and specific computation capabilities. From a manufacturing, efficiency and capability perspective, ASICs could significantly save on costs. The lines between Nvidia’s now diverse selection of GPUs – which are targeted toward different markets – and ASICs are blurring as customizability becomes mainstream. • FPGAs (Field Programmable Gate Arrays)* As the name suggests, FPGAs are programmable in a way that can fundamentally alter their performance, depending on the applications they are used with. Among existing processor technologies that are locked in or better for certain applications, FPGAs can be altered for multi-tasking like CPUs or parallel processing like GPUs. This flexibility, when delivered at scale, can be tremendously useful.

*FPGA is an integrated circuit designed to be configured by the customer or a designer after manufacturing. It features programmable logic block, and reconfigurable interconnects. That translates to flexibility in computing power. which the end customer can configure based on the application being used, wtih limitations. ◼ Total Addressable Market (TAM)20,21 The total GPU market size is expected to compound at a rate of 30.0% from 2018 to reach $103.6bn by 2026, according to a report by Acumen Research and Consulting. Nvidia presented forecasts for their potential data center and automotive markets in their Investor Day Presentation: • The Datacenter TAM is expected to reach $50.0bn in 2023 from $37.0bn in 2018 • The Auto TAM is expected to reach $25.0bn in 2025

LB•LOGIC We don’t expect the GPU market to expand at a steady compounding rate, and expect natural variations due to the cyclical nature of the industry. As a result, most semiconductor stocks experience significant volatility.

20.https://www.marketwatch.com/press-release/graphic-processing-unit-gpu-market-size-2026-by-technology-share-growth-regional- statistics-global-revenue-forecast-2018-2026-2019-09-12 21 https://investor.nvidia.com/events-and-presentations/presentations/default.aspx

“Seeking Stocks that can 11 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

◼ Quality of Product/Service Nvidia has long been regarded as the leading provider of GPUs worldwide across multiple industries. For upmarket professional applications, and especially AI and ML that utilize the CUDA platform, they are the dominant industry benchmark. Gaming enthusiasts consider AMD to be a competitive provider of GPUs at lower price points. However, Nvidia dominates the higher end market where GPU performance is given higher priority than price. ◼ History22 The company was founded in 1993 by three American computer scientists, Jensen Huang, Curtis Priem, and Christopher Malachowsky, with $40,000.0 in initial investment. In 1998, they released RIVA TNT, which solidified their reputation for building capable graphics adapters. In late 1999, they released GeForce 256, which brought on-board transformation and lighting to consumer- level 3D hardware. The GeForce outperformed existing products by a considerable margin. The success of GeForce led to the company winning a contract with Microsoft to develop the graphics hardware for the game console. In December 2000, the company reached an agreement to acquire the intellectual assets of its main rival, 3dfx, a deal that was finalized in 2002. After expansion through multiple acquisitions, Nvidia received subpoenas from the US Department of Justice regarding possible antitrust violations, as did its competitor, AMD. In 2006, the company unveiled CUDA, their proprietary architecture for general-purpose GPU (or GPGPU) applications, targeting scientists and researchers keen to harness parallel computing capabilities. Nvidia was named Company of the Year by Forbes in 2007. In 2010, GPUs powered the world’s fastest supercomputer in China. The company proceeded to expand across mobile processors, releasing DRIVE and Jetson in 2015, DGX in 2016 and the AI-focused Volta architecture in 2017. In 2019, Nvidia acquired Mellanox for $6.9bn, a leader in high performance interconnect technology. ◼ Founder/Current Management Jensen Huang – Founder, President and CEO23 • Founded Nvidia in 1993 and has served as CEO, President, and Board member since • Awards/Honours o Recipient of the IEEE Founder’s Medal o Dr. Morris Chang Exemplary Leadership Award o Honorary Doctorate Degrees from Taiwan’s Chiao Tung University and Oregon State University o In 2019, named by Harvard Business Review as #1 on their list of world’s 100 best-performing CEOs over their lifetime tenure o In 2017, named Fortune’s Businessperson of the Year

22 https://www.nvidia.com/en-us/about-nvidia/corporate-timeline/ 23 https://nvidianews.nvidia.com/bios/jensen-huang

“Seeking Stocks that can 12 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

• Previously worked at LSI Logic and AMD before founding Nvidia • Education: o MS Electrical Engineering from Stanford University o BS Electrical Engineering from Oregon State University

Chris A. Malachowsky – Founder, Nvidia Fellow24 • Founded Nvidia in 1993 and serves as a member of the executive staff and a senior technology executive for the company o Led numerous functions, including IT, Operations and all facets of the company’s product engineering o Responsible for Nvidia’s research organization, which is tasked with developing strategic technologies to drive the company’s future growth and success • Prior to Nvidia, held engineering and technical leadership positions at HP and Sun Microsystems • Holds close to 40 ; recognized authority on integrated circuit design and methodology • Education: o MS Computer Science from Santa Clara University; Distinguished Alumnus o BS Electrical Engineering from University of Florida; Distinguished Alumnus • Board member of the Computer History Museum, Hiller Aviation Museum, and LA County Museum of Art’s Art & Technology Lab • Received an Emmy award for a film he helped produce that won Best Documentary in 2009 ◼ Capital Allocation (M&A, buybacks, expansion) Acquisitions • March 12, 2019: Announced the acquisition of for $6.9bn.25 Mellanox is a leading company in high-performance computing (HPC). Together with Nvidia’s platform, they’d interconnect power to over 250 of the world’s top 500 • July 23, 2013: Acquired The Portland Group26, a producer of commercially available Fortran, C, and C++ compilers • Other notable acquisitions include , Hybrid Graphics, ULI Electronics, Exluna from 2000-2010

LB•LOGIC The Mellanox acquisition, which was recently cleared by regulations, is one of the largest acquisitions in the company’s history. We view it favorably as a strategic move to strengthen competitive positioning in HPC, which will help expand the data center business.

24 https://nvidianews.nvidia.com/bios/chris-a-malachowsky 25 https://nvidianews.nvidia.com/news/nvidia-to-acquire-mellanox-for-6-9-billion 26 https://insidehpc.com/2013/07/nvidia-acquires-the-portland-group-to-double-down-on-hpc/

“Seeking Stocks that can 13 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

Financing activities • 2017: Issued $1. 0bn of 2.20% notes due 2021 and $1.0bn of 3.2% notes due 2026 • January 22, 1999: Nvidia went public trading as NVDA on Nasdaq. The company issued 3.5mn shares for $12.00 per share; shares surged 64.0% on the first day of trading, closing at $19.69, giving the company a market value of $626.1mn

As of October 2019, the outstanding payable of 2021 and 2026 notes had a carrying value of $1.99bn

◼ Insider Transactions27 Insiders bought 1,735,160 shares and sold 2,197,437 shares in the last twelve months. • Insiders bought 101,676 shares and sold 249,480 shares in the last three months

27 https://www.nasdaq.com/market-activity/stocks/nvda/insider-activity

“Seeking Stocks that can 14 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

 TREND ANALYSIS ◼ Income Statement $ in millions FY 2015 FY 2016 FY 2017 FY 2018 FY 2019 Net Revenue 4,682 5,010 6,910 9,714 11,716 Non-GAAP Gross Profit 2,611 2,846 4,088 5,844 7,233 Non-GAAP Gross Margin 55.8% 56.8% 59.2% 60.2% 61.7% Non-GAAP Operating Income 954 1,125 2,221 3,617 4,407 Non-GAAP Operating Margin 20.4% 22.5% 32.1% 37.2% 37.6% Non-GAAP Net Income 801 929 1,851 3,085 4,143 Non-GAAP Net Margin 17.1% 18.5% 26.8% 31.8% 35.4% Non-GAAP EPS ($) 1.42 1.67 3.06 4.92 6.64

◼ Balance Sheet $ in millions FY 2015 FY 2016 FY 2017 FY 2018 FY 2019 Cash 7,201 7,370 9,841 11,241 13,292 Total Assets 2,783 2,901 4,079 3,770 3,950 Total Liabilities 497 596 1,766 4,002 782

◼ Cash Flow $ in millions FY 2015 FY 2016 FY 2017 FY 2018 FY 2019 CFO 906 1,175 1,672 3,502 3,743 CFI -727 -400 -793 1,278 -4,097 CFF -834 -676 291 -2,544 -2,866

FCF 783 1,089 1,496 2,909 3,143 FCF Margin 16.7% 21.7% 21.6% 29.9% 26.8%

LB•LOGIC The company has excellent profitability margins and is strongly FCF generative. We believe they’re well positioned to withstand down cycles in the industry and economy, while reinvesting in their considerable growth opportunities.

“Seeking Stocks that can 15 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

◼ Valuation 18.00x EV/LTM Sales 16.00x 14.00x 13.80x 12.00x 10.00x 11.21x 8.00x 6.00x Average 4.00x Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

13.00x EV/NTM Sales

11.00x 10.81x

9.00x 9.34x 7.00x

5.00x Average 3.00x Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

60.0x EV/LTM EBITDA 55.0x 50.0x

40.0x

30.0x 33.8x

20.0x Average 10.0x Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

“Seeking Stocks that can 16 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

40.0x EV/NTM EBITDA 35.0x

30.0x 28.4x 25.0x 25.8x 20.0x

15.0x Average 10.0x Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

60.0x EV/LTM FCF

50.0x

40.0x 37.5x 37.2x 30.0x

20.0x Average 10.0x Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

LB•LOGIC NVDA is trading at a Fwd EV/EBITDA of 28.4x. We believe this valuation presents further upside backed by strong growth in the datacenter market. Upcoming 7nm architecture releases will likely spur some sales growth in 2020.

“Seeking Stocks that can 17 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

◼ Recent Price Action The share price has experienced significant volatility over the past two years. Growth was inorganically boosted by the cryptocurrency bubble in 2017, leading to demand surges in gaming GPUs. As mining became less profitable, and alternatives to GPUs such as crypto-specific ASICs entered the market, demand slowed. Factored in with trade tensions and systemic drawdowns across the US market, September 2018 to December 2018 saw Nvidia lose nearly half its market value. The share price has since recovered and has surged upward, likely due to trade deal optimism and secular demand in non-gaming applications.

$350.00 Share Price Movement $300.00 $235.46 $250.00 $200.00 $150.00 $180.12 $100.00 $50.00 200 days MAVG - Dec-17 Mar-18 Jun-18 Sep-18 Dec-18 Mar-19 Jun-19 Sep-19 Dec-19

 QUALITATIVE INFORMATION ◼ Special/Unique Characteristics of the Company Market Leader Advantages Nvidia benefits from economies of scale due to its large market share and sales volumes. Significant R&D investments result in products that can scale across far more customer classes than their competitors, and this positions them extremely well from a financial perspective. CUDA, Software, and Developer Communities CUDA is a foundational platform and API that is now widely adopted by developers and end markets for AI capabilities and development. There are other open-source platforms, but the strength of the hardware and software optimization Nvidia provides gives its ecosystem a clear edge that AMD cannot compete with in developer-heavy markets.

LB•LOGIC While upgrading from a Nvidia GPU to an AMD GPU that offers a better price to performance ratio is an easy decision, doing the same if your professional work relies on CUDA makes the transition difficult. If your development relies on interaction and teamwork with others that use CUDA, your only option is Nvidia.

“Seeking Stocks that can 18 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

◼ Catalysts Tech advances Nvidia has kept up with Moore’s law for processor performance and outpaced it. A technological breakthrough may lead to catalyzed business growth as the end-users of GPUs and SOCs, particularly businesses that rely on performance to compete (compute server-side hosts such as AWS, GCP, Azure) will need to upgrade their GPU stacks more frequently. Use case demand Autonomous vehicles haven’t gone into the mainstream. When Level 5 autonomy is achieved amongst vehicle manufacturers, the SOC market for auto chips could catalyze growth and transform the industry. Nvidia appears to be well positioned to grow within the auto market when adoption occurs. Similarly, IoT, Virtual Reality are avenues of potential catalyzed expansion for parallel computing demand. ◼ What We Expect Our outlook on Nvidia depends on the balance between reduced sales growth in gaming GPUs and market opportunities for the non-gaming segments. Decreased sales growth in 2019 doesn’t concern us due to the cyclical nature of GPU demand. Furthermore, the company has excellent margins and the ability to generate free cash flows and profitability, which shall defend against down-cycles. Short term negative surprises in GPU sales in the coming quarters may lead to some share price drawdowns and volatility, while macroeconomic events would have impacts across the semiconductor industry. Regardless, the competitive moat and the technological edge Nvidia has over the competition gives us high confidence in their ability to remain the top GPU manufacturer over the coming years. AMD spent several years moving to 7nm chips across their CPUs and GPUs; they managed to challenge Intel but have fallen slightly short of being able to meaningfully challenge Nvidia’s current 12nm chips, which we expect to be replaced by 7nm architectures shortly. The rise of Google’s TPUs, attempts at GPUs by other big-tech firms and expansions of ASICs may be threats in the long term but odds for Nvidia’s continued dominance are likely heading into 2020. The R&D cycles for microprocessor development are notoriously long and difficult. More funding cannot accelerate R&D. We are optimistic about the continued rise in high-performance computing, data center applications, and autonomous vehicles. When autonomous vehicle GPU applications do accelerate, they will have industry-transforming changes. The timing is difficult to predict and depends on a variety of factors, but Nvidia appears to be strongly positioned to capitalize on the expansion when it finally occurs, given its partnerships with major auto manufacturers.

“Seeking Stocks that can 19 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

 RISKS TO THE BUSINESS Competition AMD’s transformation has been impressive over the past few years, and they may be able to challenge Nvidia in the short run. Likewise, Google or Amazon may release capable chips faster than anticipated, leading to a decline in data center growth for Nvidia. If Google attempts to develop TPUs and markets them outside the GCP ecosystem, they may take away market share from Nvidia. The chip industry is moving from “one chip fits all” to “a chip for every application.” ASICs and FPGAs are thus gaining traction and are alternative approaches that can meaningfully challenge Nvidia if competitors do make some technological leaps. Recession Risks The semiconductor industry performs notoriously poorly in recessions as consumer spending evaporates. Nvidia’s business can incur stress in a downturn. Geopolitical Nvidia has a globally diversified supply chain. Trade tensions and trade deals have impacts across the company’s operations. Increasing trade tensions may lead to surges in operational costs and can directly impact the bottom line. Key-man Risk Jensen (Jen Hsun) Huang is the visionary founder of Nvidia and has led the company as CEO to become the dominant GPU player globally with a $140.0bn market cap. His absence could severely impact the company.

“Seeking Stocks that can 20 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

 VISUAL REPRESENTATION OF CAPITAL STRUCTURE Percentage of Enterprise Value

Revolving Credit 0.42% Facility $575.0mn

1.45% 2021 and 2026 Notes $1.99bn

1.86% Total Debt $2.55bn

105.3% Market Capitalization $144.1bn

“Seeking Stocks that can 21 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

 GLOSSARY EBITDA: Earnings Before Interest, Taxes, Depreciation, and Amortization. A measure of a company’s operating performance, which allows the investor to analyze the earning power of an enterprise without having to consider financing or accounting decisions or tax environments.

EV/EBITDA: A valuation metric intended to describe the total amount an acquirer would have to pay to purchase a company (incorporating the cost of assumed debt) per dollar of EBITDA.

Fwd: Shorthand for Forward. Generally means “Next Twelve Months”.

Market Cap: Market Capitalization. Calculated by multiplying stock price by tonumber of outstanding shares.

ROA: Return on Assets. A measure of financial performance that shows the percentage of profit that a company earns in relation to all of its available resources (both debt and equity). Calculated by dividing Net Income by Average Total Assets.

ROE: Return on Equity. A measure of financial performance that shows the percentage of profit that a company earns for each dollar of shareholder equity. Calculated by dividing Net Income by Average Shareholder Equity.

ROIC: Return on Invested Capital. A measure of financial performance intended to evaluate a company’s growth and measure how efficiently a company is using investor funds to generate income. An ROIC of 2% or more in excess of a company’s cost of capital defines a value creator. Calculated by dividing after-tax Operating Income by the book value of all Invested Capital

TTM: Shorthand for “Trailing Twelve Months”.

YTM: Shorthand for “Yield to Maturity”. YTM is the total return expected if an investor holds a bond to maturity, with the assumption that all coupon and interest payments are made on schedule.

GPU: A is a computer chip that performs rapid mathematical calculations, primarily for the purpose of rendering images.

CUDA: Compute Unified Device Architecture. CUDA® is a parallel computing platform and programming model developed by NVIDIA for general computing on graphical processing units (GPUs).

CPU: . The CPU is the primary component of a computer that processes instructions. It runs the operating system and applications, constantly receiving input from the user or active software programs. It processes the data and produces output, which may stored by an application or displayed on the screen.

ASIC: Application-Specific Integrated Circuit. An application-specific integrated circuit is an integrated circuit (IC) chip customized for a particular use, rather than intended for general-purpose use.

FPGA: A Field-Programmable Gate Array is an integrated circuit designed to be configured by a customer or a designer after manufacturing

HPC: High-Performance Computing is the use of parallel processing for running advanced application programs efficiently, reliably and quickly. The term applies especially to systems that function above a teraflop or 1012 floating- point operations per second.

AI: Artificial Intelligence. The theory and development of computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.

ML: Machine Learning is an application of artificial intelligence (AI) that provides systems the ability to automatically learn and improve from experience without being explicitly programmed.

“Seeking Stocks that can 22 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

DL: Deep learning is a subset of machine learning in artificial intelligence (AI) that has networks capable of learning unsupervised from data that is unstructured or unlabeled.

“Seeking Stocks that can 23 Double in 2-3 Years”

0.94 EQUITY RESEARCH REPORT | 12/20/19

DISCLAIMER: This Report is provided for informational purposes only and is prepared without regard to the investment objectives, financial situation, or needs of any investor. The Report is not intended, and should not be relied upon, as a source of any investment recommendation, makes no implied or express recommendation to hold, sell, purchase or take any other action with regard to a security, and is not an offer or solicitation for the purchase or sale of the security that is the subject of the Report. must exercise their own independent judgment as to the suitability of a security.

Past performance is not indicative of future performance. The price of securities can and will fluctuate, and any individual security may become worthless. A high or favorable rating, rating outlook, gauge, or similar opinion is not indicative of future performance, and no user should rely on any such rating, rating outlook, gauge, or similar opinion to predict performance or potential for return. Future performance may not equal projected or forecasted performance or potential for return. All ratings and related analysis, as well as data, statistics, analysis and opinions contained herein are solely statements of opinion, and are not statements of fact or recommendations to purchase, hold, or sell any security or make any other investment decisions.

THE REPORT IS PROVIDED ON AN "AS IS" AND "AS AVAILABLE" BASIS WITHOUT REPRESENTATION OR WARRANTY OF ANY KIND. LEFT BRAIN INVESTMENT RESEARCH LLC DISCLAIMS ALL EXPRESS AND IMPLIED WARRANTIES WITH RESPECT TO THE REPORT, INCLUDING, BUT NOT LIMITED TO, ANY WARRANTIES OF ACCURACY, TIMELINESS, COMPLETENESS, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE OR USE.

The Report is current only as of the date set forth herein. Left Brain Investment Research LLC (LBIR) has no obligation to update the Report or any material or content set forth herein.

LBIR is an affiliate of Left Brain Wealth Management LLC, an investment advisor registered with the Securities and Exchange Commission. LBIR is an affiliate of Left Brain Capital Appreciation Fund, L.P., Left Brain Capital Appreciation Offshore Ltd, and Left Brain Capital Appreciation Master Fund, Ltd., all of which are hedge funds managed by Left Brain Capital Management, LLC. The general partner of these hedge funds, Left Brain Capital Management, LLC, is an affiliate of LBIR.

© 2019, Left Brain Investment Research LLC. All rights reserved. Reproduction in any form is prohibited.

“Seeking Stocks that can 24 Double in 2-3 Years”