🤖 NVIDIA: The iPhone moment of AI
Software differentiation could make a big difference in the Generative AI wave
Greetings from San Francisco! 👋
Welcome to the new members who have joined us this week!
Join the 33,000+ How They Make Money subscribers receiving insights on business and investing every week.
This article is brought to you by App Economy Portfolio.
Did you know I run one of the most popular premium investment research services on Seeking Alpha?
It’s called App Economy Portfolio. That’s where I share my entire portfolio and trades in real-time (what I’m buying, selling, and why).
You’ll get the following:
Actionable stock ideas (deep dives).
Timely lists of stocks that are at the top of my watch list.
Live trade alerts (I invest a fixed amount monthly spread across 4 to 5 stocks).
I cover the earnings of 70+ companies in my portfolio every quarter and update my ratings accordingly in our private community chat.
Want to advertise in How They Make Money? Book here.
“We are at the iPhone moment of AI.”
These are the words of NVIDIA CEO Jensen Huang during the latest semi-annual GTC (GPU Technology Conference) a few days ago.
In case you missed it, we discussed how NVIDIA makes money back in December 2022. Since then, the stock is up a whopping 60%. 👀
With the release of its Q4 FY23 earnings report (ending January 2023) and the unveiling of several new products at the GTC event, it's time for another update on the ever-evolving world of AI and NVIDIA's role in it.
Today, we’ll cover the following:
NVIDIA Q4 FY23.
Recent business highlights.
Key quotes from the earnings call.
What to watch looking forward.
If you are new to the semiconductor industry or find terms like CPU, GPU, and TPU puzzling, don't worry, we've got your back!
We just released a new video on the How They Make Money YouTube Channel where we dive into the bottom of the AI tech stack and explain the roles of companies like NVIDIA, AMD, and TSMC in the AI race.
The semiconductor industry is very cyclical. High demand can lead to supply shortages, followed by gluts. In addition, the market has been impacted by COVID-related disruptions.
Shortages, gluts, and pandemics have one thing in common: they are temporary by nature. So judging NVIDIA only by its financials in the past 12 months would be a mistake. At the end of this article, we'll discuss its current valuation and what it means for investors.
Now, let's dive into the most recent quarter!
1. NVIDIA Q4 FY23
Here is a bird’s-eye view of the income statement.
I’m focusing on sequential growth (Q/Q) to show how things are improving (or not) following the past two challenging quarters.
Note that 69% of NVIDIA’s revenue came from outside the US in FY23.
Revenue grew +2% Q/Q (and declined -21% Y/Y) to $6.1 billion ($30 million beat).
⚙️ Data Center declined by -6% Q/Q to $3.6 billion (or +11% Y/Y).
🎮 Gaming grew +16% Q/Q to $1.8 billion (or -46% Y/Y).
👁️ Professional Visualization grew +13% Q/Q to $0.2 billion (or -65% Y/Y).
🚘 Automotive grew +17% Q/Q to $0.3 billion (or +135% Y/Y).
🏭 OEM & Other grew +15% Q/Q to $0.1 billion (or -56% Y/Y).
The graph below shows the past nine quarters, which paint a challenging time in FY23 caused by the decline in the gaming segment.
Gross margin 63% (+10pp Q/Q), vs. a 63% guidance.
Operating margin 21% (+11pp Q/Q).
EPS was $0.88 ($0.08 beat).
Operating cash flow was $2.25 billion (37% margin, -3pp Y/Y).
Free cash flow was $1.7 billion (29% margin, +31pp Y/Y).
Cash and cash equivalent: $13.3 billion.
Debt: $11.0 billion.
Q1 FY24 Guidance:
Revenue to grow +7% Q/Q (decline -22% Y/Y) to $6.5 billion ($0.18 billion beat).
Gross margin 64% (+1pp Q/Q).
So what to make of all this?
Despite a challenging environment, the quarter wasn't as bad as anticipated. Revenue slightly exceeded modest expectations.
Data Center remains the focus, and while it declined 6% sequentially due to temporary headwinds in China, it still made up 60% of Q4 revenue.
Gaming bounced back after a post-COVID slump. The channel inventory challenges may be behind us.
Professional Visualization and Automotive also saw double-digit sequential growth. In addition, the "OEM & Other" category recovered after decreased demand for mining due to the crypto winter.
Margins improved across the board. Following inventory improvements, the gross margin was in line with guidance (+10pp Q/Q). This led to an increased operating margin (+11pp Q/Q).
Operating expenses grew +27% Y/Y, mainly due to higher compensation expenses. In short, the company invested in R&D as revenue declined, causing a Y/Y shrink in operating margin. Note that operating expenses were flat Q/Q.
Q1 FY24 revenue guidance surpassed Wall Street's expectations by 3%, with a stronger rebound than anticipated.
Management expects profitability to increase from here. The Q1 FY24 outlook includes a GAAP operating margin of 25% (+4pp Q/Q)
Overall, if we look at the full year FY23, revenue remained flat and the operating margin hit a low similar to FY20 (see graph).
Is the business sustainable?
NVIDIA has a positive net cash position and maintained a positive operating cash flow ($5.6 billion in FY23) through a challenging part of the cycle. Capital expenditures expected in FY24 ($1.2 billion) are relatively small.
Management repurchased $1.2 billion of NVDA shares in the past three months and $10 billion in the past 12 months. There’s still $7 billion left in the current share repurchase program, lasting through December 2023.
Considering the stock's recent performance, it was a great use of capital (though the jury is still out).
2. Recent business highlights.
New announcements at GTC 2023
NVIDIA unveiled many new initiatives around software and services at its GTC 2023 keynote, mainly focusing on AI. Here's what we learned:
NVIDIA DGX Cloud.
Dubbed "an AI Factory in the Cloud," it's the world's first AI supercomputer in the cloud, offering a multi-node AI-training-as-a-service solution tailored for enterprise AI demands.
NVIDIA AI Foundations.
NVIDIA announced new cloud services for language, visual media, and biology, enabling companies like Adobe to create, refine, and operate custom generative AI and large language models (LLM). It includes:
NVIDIA NeMo (language): A toolkit for building conversational AI models.
NVIDIA Picasso (visuals): Generates photorealistic images, video, and 3D.
NVIDIA BioNeMo (biology): Pre-trained and 3D protein prediction models.
NVIDIA launched four inference platforms for LLM and Generative AI workloads (chatbots, text-to-image content, AI video, and more). Google Cloud is a key cloud partner and early customer.
Partnership with Microsoft to connect Omniverse Platform to Microsoft 365 applications.
We previously discussed Omniverse, a platform for industrial visualization and 3D design. Microsoft will bring the industrial Metaverse to millions of existing Enterprise Users through Azure Cloud.
NVIDIA brings accelerated computing to computational lithography.
A breakthrough allows ASML, TSMC, and Synopsys to speed up the design and manufacturing of next-gen chips. It's called the NVIDIA cuLitho software library and is being integrated by market leaders.
The keynote is worth a watch if you want to feel inspired and energized by the upcoming AI innovations.
3. Key quotes from the earnings call
CFO Colette Kress touched on Data Center (-6% Q/Q and +11% Y/Y):
“[…] some cloud service providers paused at the end of the year to recalibrate their build plans. Though we generally see tightening that reflects overall macroeconomic uncertainty, we believe this is a timing issue as the end market demand for GPUs and AI infrastructure is strong. […] The total data center sequential revenue decline was driven by lower sales in China, which was largely in line with our expectations, reflecting COVID and other domestic issues.
[…] We expect a great year with our year-over-year growth in data center probably accelerating past Q1.”
About the new flagship H100 data center GPU:
“Adoption of our new flagship H100 center GPU is strong. In just the second quarter of its ramp, H100 revenue was already much higher than that of A100, which declined sequentially. This is a testament of the exceptional performance on the H100, which is as much as nine x faster than the A100 for training and up 30x cluster in inferencing of transformer-based large language models.”
“Sequential growth was driven by the strong reception of our 40 Series GeForce RTX GPUs based on the Ada Lovelace architecture.
The year-on-year decline reflects the impact of channel inventory correction, which is largely behind us.
[…] The flagship RTX 4090 has quickly shot up in popularity on Steam to claim the top spot for the AI architecture, reflecting gamers' desire for high-performance graphics.
Earlier this month, the first phase of gaming laptops based on the Ada architecture reached retail shelves, delivering NVIDIA's largest-ever generational leap in performance and power efficiency.”
On GeForce NOW cloud gaming:
“It now has more than 25 million members in over 100 countries. […] We agreed to a 10-year partnership to bring to GeForce NOW Microsoft's lineup of Xbox PC games”
On Pro Visualization:
“Sequential growth was driven by desktop workstations with strengths in the automotive and manufacturing industrial verticals. Year-on-year decline reflects the impact of the channel inventory correction, which we expect to end in the first half of the year.
Interest in NVIDIA's Omniverse continues to build with almost 300,000 downloads so far, 185 connectors to third-party design applications.”
“Sequential growth was driven primarily by AI automotive solutions. New program ramps at both electric vehicle and traditional OEM customers helped drive this growth.
[…] The NVIDIA Drive operating system received safety certification from TÜV SÜD, one of the most experienced and rigorous assessment bodies in the automotive industry.”
CEO Jensen Huang on Generative AI and Hopper:
“Generative AI's versatility and capability has triggered a sense of urgency at enterprises around the world to develop and deploy AI strategies. Yet, the AI supercomputer infrastructure, model algorithms, data processing and training techniques remain an insurmountable obstacle for most.
[…] We're introducing Hopper, which, with its transformer engine, it's new NVLink switches and its new InfiniBand 400 gigabits per second data rates, we're able to take another leap in the processing of large language models.”
On NVIDIA DGX Cloud:
“AI supercomputers are hard and time-consuming to build. Today, we are announcing the NVIDIA DGX Cloud, the fastest and easiest way to have your own DGX AI supercomputer, just open your browser. NVIDIA DGX Cloud is already available through Oracle Cloud Infrastructure and Microsoft Azure, Google GCP and others on the way.
[…] we're effectively attracting customers to the CSPs (cloud service providers). This is a very, very exciting model for them. And they welcomed us with open arms. And we're going to be the best AI salespeople for the world's clouds.”
During the Q&A, Jensen touched on the monetization of the AI software suite after Colette Kress pointed out that software revenue was already in the “hundreds of millions.”
“NVIDIA AI is essentially the operating system of AI systems today. […] It accelerates people's adoption of this machine learning pipeline, which is not for the faint of heart. It is a very extensive body of software. It is not deployed in enterprises broadly, but we believe that by hosting everything in the cloud, from the infrastructure through the operating system software, all the way through pretrained models, we can accelerate the adoption of generative AI in enterprises.”
He beautifully explained the revolution triggered by ChatGPT:
“The world now realizes that maybe human language is a perfectly good computer programming language, and that we've democratized computer programming for everyone, almost anyone who could explain in human language a particular task to be performed.”
4. What to watch looking forward
As the global pandemic intensified the need for at-home entertainment, a combination of skyrocketing demand and limited supply caused a chip shortage in 2021. NVIDIA's gaming segment saw reduced shipments in 2022 as the company worked to align channel inventory (products already on shelves).
As a result, gaming revenue dropped dramatically in the past three quarters. However, now that inventory levels have stabilized, the gaming segment may have reached its low point in Q3 and could see a continued rebound in the upcoming quarters after growing +16% Q/Q in Q4 FY23.
We previously discussed the successful launch of the Ada architecture, designed to deliver exceptional AI and compute performance for gaming and creative work. In addition, the flagship RTX 4090 has quickly gained popularity on Steam since its launch in October 2022.
As a gaming industry veteran, I expect gaming to benefit from long-term tailwinds well beyond the post-COVID hangover. Each new generation shows higher gaming penetration (see chart). PC gaming might become a smaller portion of the market, but the overall industry is expected to expand with new needs around streaming.
A word on valuation
NVDA has become a polarizing stock due to its popularity among retail investors and high valuation.
To make matters worse, the valuation multiples based on backward-looking metrics are inflated due to the recent underperformance of the gaming segment. Since earnings (the denominator) are temporarily weaker, the P/E ratio becomes temporarily inflated.
As is often the case, zooming out is the most critical step to gain context.
Historically, NVIDIA beats Wall Street’s EPS expectations by a wide margin (with the notable exception of Q2 and Q3 FY23).
Even assuming the company merely meets expectations moving forward, the stock is trading at:
36 times FY26 EPS (January 2026, 3 years from now).
22 times FY28 EPS (January 2028, 5 years from now).
Given the track record of outperformance demonstrated by this management team over the years, the company may be less overvalued than it seems at first glance.
Of course, based on my Twitter feed, taking the 30,000-foot view isn't the area of expertise of opinionated anonymous accounts.
The AI supercycle
NVIDIA has touted a $1 trillion TAM (total addressable market).
During the latest earnings call Q&A, Jensen Huang came short of saying the opportunity is even larger now after the emergence of ChatGPT and other Generative AI tools. But it may be attainable faster:
“I think those numbers are really good anchor still. The difference is because of the, if you will, incredible capabilities and versatility of generative AI and all of the converging breakthroughs that happened towards the middle and the end of last year, we're probably going to arrive at that TAM sooner than later.”
NVIDIA made $27 billion in revenue in FY23 but still has a vast market to tackle. Rising competitors and new developments will define this unique “iPhone moment in AI.”
So buckle up, and I hope you enjoy the ride as much as I do.
That’s it for today!
Stay healthy and invest on!
Sign up for free to How They Make Money so you don’t miss any updates!
Disclosure: I am long NVDA in the App Economy Portfolio, though I haven’t added to my position since July 2022. I share my ratings (BUY, SELL, or HOLD) with App Economy Portfolio members here.
Thanks again for another great article!🙏🙏🙏