Long Nvidia
Long-term strategy
Price: Buy when NVDA PE drops to 40ish; NVDA PE average/median PE is 58 in the past 5 years, lowest: 16.9; highest: 100 - double check
Portfolio position: Bet on AI in the next 20 years - NVDA should be one of our major holdings (15% to 20% of our portfolio), we should build our position in the next 1-3 years. Be patient.
Lesson learnt: Do not clean up the entire position if we want to own the company for the long-term, especially considering growth stocks (TSLA and NVDA are volatile).
Short-term strategy: Buy NVDA and PANW a month before earnings (quiet period) and sell afterwards just for short-term gain!
Stock rises after earning release regardless - Jensen and Nikesh know how to pop the stock!
Look at the price actions after NVDIA’s earnings calls, Jensen and CFO apparently knows how to manage Street expectations.
My revenue prediction was correct - no high growth on both segments (graphic + computer/networking) and the outlook was not great, but stocks popped regardless - with slightly beaten the earning expectations from the street and delivered what the street wanted to hear - AI was mentioned 73 times in the earnings call.
Valuation
As of 2/23/2023
Key Metrics to Monitor
Revenue growth - should above 50% overall if we are targeting PE at 40ish
Revenue mix - how data center, gaming, and autonomous driving increases.
Vertical expansion across industry - currently in CSP, Gaming, Auto (Mercedes is the first one), Bank (Deutsche bank is the first one), Biology (NVDA expands large language model to biology), and what else?
Adoptions and revenue growth of Omniverse, CUDA users?
Operating Margin - should always go beyond 65% to be considered as a SW company.
Debt to equity ratio - currently 11B in long-term debt with 13 B cash on hand.
Leadership and Management
Jensen Huang
Co-founder
Boldly visionary - can tell a story that Wallstreet can understand and love.
Proven ability to take the company to the next level a few times.
Long-term commitment - 60 years old by 2023.
Senior leadership with proven skills
Flat organization design - Build a company / organization that can pivot and change course fast.
Speed and adaptability to the changing world is important, and the great company we want to invest in has one thing in common - a flat organization with no BU structure: TSLA, META, GOOG, Amzn, NVDA.
Business
New update from 10-K (FY23 Annual Report) - end Jan'23
Overall - dded GPU-powered deep learnings (recommendation, decision-making), large language models, and generative AI to be part of the business.
It’s supercomputer is now covering 23 of the top 30 systems on the Green500 list demonstrating its strength in energy sufficiency and the choice of ESG.
Data Center becomes its top revenue generation business (was Gaming in FY22 and before)
New offering
CUDA-X programming model, SW libraries and SDKs, application framework and service as part of the platform strategy, sold through packaged with HW or separately.
DGX Cloud, a Cloud-based infrastructure + SW, with partnership with leading CSP to host these services in their data centers.
Paid licenses to NVIDIA AI Enterprise - Enterprise-grade AI SW and NVIDIA vGPU SW.
Expand Data center processor portfolio to include Bluefield DPUs (in market already) and CPUs with samples planned to ship in the first half of FY24 - to support data-center-infrastructure-on-a-chip SW, or DOCA to enable developers build SW-defined, HW-accelerated networking, security, storage, and management applications for BlueField DPUs.
Ada Lovelace Architecture - RTX on gaming with 4x performance than the previous generation.
Automotive - In addition to AV, it will provide infotainment platform solutions too. It is going to be a 50-50 revenue split with the car manufactures.
Concentration of revenue - Non U.S revenue is now 69% vs. 84% in 2022. The decline in revenue outside of the U.S was primarily driven by China and Taiwan related to Data Center and Gaming.
Tax - 4.5% increased from 1.9% in FY22.
$7.23 billion remaining under share repurchase till Dec, 2023.
GPU architecture/platforms for scientific computing, artificial intelligence, data science, autonomous vehicles, robotics, and augmented and virtual reality, AR and VR.
Powers over 70%, and 8 of the top 10, supercomputers on the global TOP 500 list.
Report its business in two segments:
Graphics segment
GeForce GPUs for gaming and PC.
GeForce NOW game streaming services and membership
RTX GPUs for enterprise workstation graphics
vGPU, or virtual GPU, SW for Cloud-based visual computing
Omniverse SW for 3D designs and virtual worlds
Compute & Networking segment for Data Center platforms and systems for AI, HPC, and accelerated computing.
Mellanox networking and interconnect solutions
Automotive AI Cockpit
Autonomous vehicle solutions
CMP - Cryptocurrency mining processors
Addressable Markets
NVIDIA addresses four large markets: Gaming, Data Center, Professional Visualization, and Automotive.
According to NVIDIA (in investor day), its TAM is 300 billion HW plus 300 billion SW.
Strategies
Advancing the accelerated computing platform - a unified underlying architecture leveraging GPUs and CUDA as the fundamental building blocks with programmable nature.
Lead AI in both training and inference - this includes GPUs, interconnects, systems, CUDA programming language, algorithms, libraries, and tooling.
Apple isn’t the only company that understands the power of differentiating premium hardware with software - cuda is free (SDK, and toolkits). CUDA only works with Nvidia video cards, in large part because many of the routines are hand-tuned and optimized.
Extending leadership in computer graphics - Omniverse is the real-time 3D design collaborating and virtual world (digital twin) that is the engine, the only Universal Graphic database in the Cloud.
Lead autonomous vehicle platform - provide a full functionally safe AI-based hardware and software solution for the AV market under DRIVE to enable AI for autonomous driving - perception, localization, and planning.
License IP - Licensing and development agreement with customers and partners.
NVIDIA now has a platform strategy as well as AI-Model-as-a-service new business model to be announced at its annual GTC, March 20-21, 2023. New business model basically has three parts.
There’s the lowest-level part, which is all the Ada Lovelace GPUs and the Omniverse computers, and some of it will be in the cloud, some of it will be on-prem, and some of it will be at the far edge. A long time ago, and Akamai probably commercialized the best, came up with the content delivery network, the CDN. In the future, there’ll be a GDN (graphics delivery network), and the GDN is a whole bunch of GPUs very far edge, and it makes it possible for content to be enjoyed interactively and so low-latency that you can actually use VR and AR in the cloud very, very easily. It’ll track your head pose and it’ll track the world perfectly, and the round trip will be in not tens of milliseconds, but it’ll probably be ten milliseconds. So in that future world the metaverse could manifest wherever and be rendered beautifully right there on the spot. So on the lowest level, it’s the graphics hardware, it’s the infrastructure, and it’s in the clouds and the prem and such.
Then the next layer is the database, and the business model there is probably going to be very Snowflake-y, if you will. You could host your own database, we’re going to make it possible for you to host your own database, in which case you could imagine that’s probably very Oracle or very SAP-like, except for the metaverse. It could be in the cloud, it could be very Snowflake-like.
Then there’s the viewports, and my sense is that for the viewports, there’ll be two different types of viewports. For most of our consumer application viewport providers, they could be a social network, they could be a game developer, we’ll just let them do that for free. Then for the industrial applications, like Digital Twins, we might create something like DRIVE Sim or Isaac SIM that has a lot of the simulation, Digital Twin nature to it, and then give that to people for them to modify, and then we’ll monetize that per engine. There’s a whole different layer to it. The one thing that we won’t do is we absolutely will not build a video game, we won’t host a social network, we won’t host a virtual world, those kinds of things. We’ll tend to probably be in the background, creating the engines.
Seasonality
Its consumer products typically see stronger revenue in the second half of its fiscal year. Some of the products typically generate stronger revenue in the second and third quarters, and weaker revenue in the fourth and first quarters.
Manufacturing
It is a fabless manufacturing strategy - use vendors for all phases of the manufacturing process, including wafer fabrication, assembly, testing, and packaging.
NVIDIA may directly procure certain raw materials used in the production (e.g. substrates and a variety of components with majority of the components procured by suppliers directly).
NVIDIA has placed non-cancellable inventory orders for certain products in advance of our historical lead times, paid premiums and provided deposits to secure future supply and capacity and may need to continue to do so in the future.
NVIDIA consign key components or materials: GPU, SoC, memory, and integrated circuit to the contract manufacturers. It uses TSMC and Samsung to produce its wafers.
Competition
AMD, Intel, and large internet services companies with internal teams designing internal solutions - Alibaba, Google, TSLA, Arista, Marvell, Cisco, Juniper, … and Amazon.
Human Capital Management
As of Jan 30, 2022. It has 22,473 employees in 32 countries with 16,242 in R&D (70%). Overall turnover rate was 4.9% in fiscal year 2022.
Concentration of Revenue
Revenue from Sales to customers outside of the U.S accounted for 84% and 81% of total revenue for fiscal years 2022 and 2021, respectively.
One customer represented 10 of the total revenue in Q4’22.
Market Indicator
Micron’s earning is 2 months ahead of NVDA
Micron’s market outlook in Dec '22 earnings was 20% down, and that's NVDA's outlook as well.
SEC
Q4’23 - 2/22/2023
Revenue growth from GPU specialized CSP customers last year significantly outpaced that of data centers as more enterprise customers moved to a Cloud-first approach.
CSP customers drove 40% of our data center revenue.
Announcing new business model - AI-model-as-a-service with NVIDIA DGX Cloud available through Oracle OCI, Azure, GCP, and others on the way.
Pricing - is priced at around $6,000 per CPU socket; is it too expensive?
NVIDIA DGX Cloud and NVIDIA’s infra can be full stack in AI start-ups around the world - they now have an instantaneous infra that is the most advanced.
7 Billion remaining for share buyback till Dec, 2023.
Total TAM - $300 billion hardware TAM and $300 billion software TAM.
Autonomous drive - Mercedes opportunity - DRIVE and Connect around %0-50 revenue split.
The management expects a great year and do expect YoY growth.
Gaming is recovering from the post-pandemic downturn.
Outlook for the first quarter of fiscal 2024 is as follows:
Revenue is expected to be $6.50 billion, plus or minus 2%.
GAAP and non-GAAP gross margins are expected to be 64.1% and 66.5%, respectively, plus or minus 50 basis points.
GAAP and non-GAAP operating expenses are expected to be approximately $2.53 billion and $1.78 billion, respectively.
GAAP and non-GAAP other income and expenses are expected to be an income of approximately $50 million, excluding gains and losses from non-affiliated investments.
GAAP and non-GAAP tax rates are expected to be 13.0%, plus or minus 1%, excluding any discrete items.
Financials
Quarterly revenue of $6.05 billion, down 21% YoY
Fiscal year revenue of $27 billion, flat from a year ago.
Quarterly and annual return to shareholders of $1.15 billion and $10.44 billion, respectively.
Pay quarterly cash dividend of 0.04 per share.
Comments
Post a Comment