Nvidia (NVDA)
Q4 Results
The hype is reality and demand is set to accelerate.
Nvidia reported excellent results which beat very elevated expectations.
We first covered Nvidia with a detailed note on 30th May 2023 and this can be found here. We followed up with two notes and they can be found here and here.
Results
Prior to these numbers the revenue situation was as shown below
For the last two quarters, demand has taken off strongly on the back of GPU demand driven by Hyperscaler Cloud players looking to boost their Generative AI capability.
Markets expected this strong demand growth to continue, and this had driven a huge move in the stock price as shown below.
Nvidia’s stock has risen 226% in a year and strongly outperformed Intel and AMD.
Results
Nvidia has an unusual year end namely January 28. The results they have just reported are for the three months, and the twelve months, to January 28, 2024. The quarter is Q4 2024.
NVDA reported revenue of $22.1bn, up 22% from the previous quarter and up 265% from a year ago.
Fully diluted EPS was $4.93, up 33% from the previous quarter and up 765% from a year ago.
For the full year to January 28, 2024, revenue was up 126% to $60.9 bn. Fully diluted EPS share was $11.93, up 586% from a year ago.
Outlook
NVIDIA’s outlook for the next quarter (Q1 25) is as follows:
Revenue is expected to be $24.0 billion, plus or minus 2%.
Gross margins are expected to be 76.3 %.
Source: Long-Term Investing
As the table above shows. the surge in revenue was driven by Datacentre demand which increased by 408% in the last year.
Datacentre now accounts for 83% of total revenue while Gaming is only 13%. Just three years ago, the numbers were 38% and 50% respectively. Nvidia is a major force in Gaming, but it is mostly now a supplier of GPUs and associated software to Datacentres and, what CEO Jensen Huang calls, Generative AI factories.
I was listening to the earnings conference late last night, London time with my son. He asked me why Nvidia does not just get out of Gaming and focus on the Datacentre opportunity only?
He may have an Nvidia GTX powered Gaming PC (which he built himself) but he is not (yet) much of an Equity Analyst.
Source: Long-Term Investing
As the Table above shows, total revenue has accelerated in the last three quarters.
Source: Long-Term Investing
The 265% growth in Revenue translated into an 808% growth in Net Profit and a 721% growth in fully diluted EPS.
Source: Long-Term Investing
As revenue has surged, margins have increased greatly due to operating leverage. In the last year operating margins have increased from 20.8% to 61.5%.
Balance Sheet
$26 bn in cash & equivalents. It had $13.3bn just 12 months ago. It has about $10 bn in debt. So net cash ~ $ 16bn.
It bought back $9.5 bn in stock during fiscal 2024.
Datacentre
The rise of generative AI and products such as ChatGPT, Copilot. LLAMA, GEMINI, and many others has led to a need for the US$ 1 trillion worth of Datacentres to upgrade their chips from CPUs to GPUs. Let us call this “upgrade demand”.
Generative AI and the investment it will demand will need another US$ 1 trillion of Datacentres in the next five year. This means there will be a doubling of Datacentre capacity in the next five years. Let us call this “new demand”.
In our view, the combination of “upgrade demand” and “new demand” will combine to great huge growth in demand for GPUs, related software, and networking products.
Nvidia has a 70% to 90% share in the supply of Gen -AI focused GPU Chipsets and while AMD (with its very good MI300 range) and others may grab some market share, this is likely to be a sellers’ market for the next three years. Nvidia has the best chips, software, and networking and, can sell as many as its supply chain (mainly TSMC) can produce, at the current high gross margins.
“Datacentre revenue for the fiscal 2024 year was $47.5 billion, more than tripling from the prior year. The world has reached the tipping point of new computing era.”
“The $1 trillion installed base of datacentre infrastructure is rapidly transitioning from general purpose to accelerated computing.”
“Companies have started to build the next generation of modern datacentres, what we refer to as AI factories, purpose built to refine raw data and produce valuable intelligence in the era of generative AI”.
Datacentres everywhere (except China) are buying the most advanced Hopper H100 and H200 based chipsets which needed demand for Training and Inference for Large Language Models (LLMS).
“Demand for Hopper remains very strong. We expect our next-generation products to be supply constrained as demand far exceeds supply. Fourth quarter datacentre growth was driven by both training and inference of generative AI and large language models across a broad set of industries, use cases and regions.”
It was recently reported that Microsoft Azure ordered 10,000 Hopper chipsets.
“In the fourth quarter, large cloud providers represented more than half of our datacentre revenue, supporting both internal workloads and external public cloud customers. Microsoft recently noted that more than 50,000 organizations (!) use GitHub Copilot business to supercharge the productivity of their developers, contributing to GitHub revenue growth accelerating to 40% year-over-year. And Copilot for Microsoft 365 adoption grew faster in its first two months than the two previous major Microsoft 365 enterprise suite releases did.”
All recommendation systems such as those used by Netflix, Meta, Amazon, and others will need to move to accelerated computing and buy new advanced GPU Chipsets
“Consumer internet companies have been early adopters of AI and represent one of our largest customer categories. Companies from search to e-commerce, social media, news and video services and entertainment are using AI for deep learning-based recommendation systems. These AI investments are generating a strong return by improving customer engagement, ad conversation and click-throughs rates. Meta in its latest quarter cited more accurate predictions and improved advertiser performance as contributing to the significant acceleration in its revenue.”
The demand is not limited to the Coud hyperscalers and big tech. Governments and other sectors are also looking to set up LLMs.
“New startups are creating LLMs to serve the specific languages, cultures, and customs of the world many regions. And others are creating foundation models to address entirely different industries like Recursion Pharmaceuticals and Generate: Biomedicines for biology. “
“One of the most notable trends over the past year is the significant adoption of AI by enterprises across the industry verticals such as automotive, healthcare, and financial services. NVIDIA offers multiple application frameworks to help companies adopt AI in vertical domains such as autonomous driving, drug discovery, low latency machine learning for fraud detection or robotics, leveraging our full stack accelerated computing platform.”
Healthcare and Financial services are likely to be big buyers.
“We have built deep domain expertise in healthcare over the past decade, creating the NVIDIA Clara healthcare platform and NVIDIA BioNeMo, a generative AI service to develop, customize and deploy AI foundation models for computer-aided drug discovery. In financial services, customers are using AI for a growing set of use cases from trading and risk management to customer service and fraud detection. For example, American Express improved fraud detection accuracy by 6% using NVIDIA AI.”
Datacentres do not just need GPUs, but they also need hugely sophisticated ways of networking the thousands of computers together. Two years ago, Nvidia acquired Mellanox which had developed the InfiniBand networking protocols and products. Microsoft Azure, for one, uses 29,000 miles of its InfiniBand connectivity cabling.
“Our Quantum InfiniBand solutions grew more than 5x year on year. NVIDIA Quantum InfiniBand is the standard for the highest performance AI-dedicated infrastructures. We are now entering the ethernet networking space with the launch of our new Spectrum-X end-to-end offering designed for an AI-optimized networking for the datacentre. Spectrum-X introduces new technologies over ethernet, that are purpose built for AI.”
Nvidia has developed DGX supercomputers which weighs 70 pounds and can have up to 35,000 parts and many Nvidia GPUs.
“DGX Cloud is used for NVIDIA's own AI R&D and custom model development as well as NVIDIA developers. We announced that NVIDIA DGX Cloud will expand its list of partners to include Amazon's AWS, joining Microsoft Azure, Google Cloud and Oracle Cloud.”
This is important as the first three are developing their own chips ad Nvidia has offered to help them with that. However, as the above indicates, they, and others, will continue to be big buyers of Nvidia chipsets as well.
“We're at the beginning of two industry-wide transitions and both of them are industry wide. The first one is a transition from general to accelerated computing.”
“.. with accelerated computing, you can dramatically improve your energy efficiency. You can dramatically improve your cost in data processing by 20 to 1.”
“That speed is so incredible that we enabled a second industry-wide transition called generative AI. Generative AI”
“You can't do generative AI on traditional general-purpose computing. You have to accelerate it.”
In addition to all the new industries governments demand is there for creating Sovereign AI.
“The reason for sovereign AI has to do with the fact that the language, the knowledge, the history, the culture of each region are different, and they own their own data. They would like to use their data, train it with to create their own digital intelligence and provision it to harness that raw material themselves. It belongs to them, each one of the regions around the world. The data belongs to them. The data is most useful to their society. And so, they want to protect the data. They want to transform it themselves, value-added transformation, into AI and provision those services themselves. So, we're seeing sovereign AI infrastructure is being built in Japan, in Canada, in France, so many other regions.”
Chip Supply:
Demand is huge but what about supply. Nvidia does not manufacture but works with partners such as TSMC and others.
“Our supply chain is just doing an incredible job for us, everything from of course the wafers, the packaging, the memories, all of the power regulators, to transceivers and networking and cables and you name it..”
The H100 chips continue to enjoy easing supply constraints. However, the new generation of chips will be supply constrained. H200 doubles inference performance compared with H100. H200 will start shipping soon but demand will exceed supply for some time.
“…and right now, we are ramping H200's. There is no way we can reasonably keep up on demand in the short term as we ramp”
Networking Revenue:
Networking crossed $10bn in annual revenue and this is likely to rise by 30% in 2025 due to demand for Spectrum X InfiniBand products.
“Spectrum-X will be our AI-optimized networking and that is ramping, and so we'll -- with all of the new products, demand is greater than supply.”
The company claims that Hopper Chipsets and Spectrum X InfiniBand are “emerging as the de facto standard for HPC and AI.”
Software:
Nvidia is also a powerhouse in software. Software is at a US $ 1bn annual revenue run rate and the company expects this to accelerate.
“All of these things require different software stacks. That's the reason why NVIDIA has hundreds of libraries. If you don't have software, you can't open new markets. If you don't have software, you can't open and enable new applications. Software is fundamentally necessary for accelerated computing.”
“my guess is that every enterprise in the world, every software enterprise company that are deploying software in all the clouds and private clouds and on-premises, will run on NVIDIA AI Enterprise, especially obviously for our GPUs. And so, this is going to likely be a very significant business over time…. it's already at $1 billion run rate and we're really just getting started.”
Gaming
The growth in Datacentre has taken the attention from Gaming. However, the company continues to progress in Gaming as well.
“NVIDIA RTX, introduced less than six years ago, is now a massive PC platform for generative AI, enjoyed by 100 million gamers (!) and creators. The year ahead will bring major new product cycles with exceptional innovations to help propel our industry forward”.
Expenses
GAAP operating expenses rose by a modest 23% Y/Y. This was due to hiring and investment in infrastructure, both needed to keep up with the soaring growth. Given triple digit revenue growth, 23% growth in OpEx is modest, and this led to the surge in operating margins noted above.
China:
The only negative point of the results is China which represented nearly 25% of total data revenue. This fell to just 5% due to the US government restrictions on the exports of the most advanced Chips. Nvidia has shifted to shipping less complex technology which is still allowed but there is a risk the Chinese suppliers will switch to chips from local suppliers, notably Huawei, and Chinese demand may never recover for Nvidia.
Summary
This was another very strong quarter. The acceleration of the last three quarters has been one of the most remarkable I have seen for any company in decades.
The company’s comments in the earnings call and the guidance suggests this demand boom is far from over. There will be some inevitable slowdown in growth in the next few years as the law of large numbers comes into play.
Valuation
The Equity analysts are likely to increase their estimates this morning in the light of these results.
Based on current estimates, Nvidia trades at a multiple of 30x FY 25 EPS. That EPS is currently expected to grow by 74% Y/Y. If this growth is achieved (and we are inclined to give management the benefit of the doubt), the stock is not currently expensive.
Conclusion
Nvidia is a remarkable company which has the best products and services in an area which is seeing greatly unprecedented demand. It has a dominant market share in an exploding market.
On August 2023 we concluded our last note on Nvidia with the following:
“The demand momentum is currently so strong that it will continue for the next year at least and there are likely to be revenue/earnings “beats” in the next two quarters as market expectations catch up with booming demand.
We will continue to hold the stock as we believe earnings growth will be strong enough to justify the current valuation. However, we are not adding to our current position. In the meantime, we will continue to track the company closely.”
We originally had a 2.0% allocation to Nvdia with the view to increasing it to 4% if the stock fell. However, we took our eye off the ball and the market never gave us the chance to increase to 4%.
Our 2.0% allocation now accounts for 7% of our portfolio. If we had originally bitten the bullet and allocated 4%, Nvidia would now be worth much more.
The lesson here is we should have remembered the words of two wise investors.
“When it rains gold, put out the bucket, not the thimble." - Warren Buffett.
“The wise ones bet heavily when the world offers them that opportunity. They bet big when they have the odds. And the rest of the time, they don't. It's just that simple.” - Charlie Munger.
We are going to increase our allocation to Nvidia today by 2%.