During the Pandemic, while some took up hobbies like baking bread and TikTok challenges, others decided to become traders. The Fed’s attempt to salvage what they could of the market, paired with easy access to trading apps like Robinhood, led to the perfect storm, ballooning the market to all-time highs during a time of great uncertainty. It was a wild time and they made a whole movie out of it, which I haven’t seen but I hear it’s good. Since then, the market has become quite sensitive, and the hype cycles move faster than ever. As I write this, the SPY is once again hitting all-time highs as we continue through this AI boom with Nvidia leading the charge. Here we go again.
Can’t Stop Won’t Stop
[Nvidia’s] Chief Executive Jensen Huang described AI as hitting “the tipping point” and indicated demand for the computing power that underlies AI remained astronomical. “Demand is surging worldwide across companies, industries and nations,” he said.
That demand showed up in the company’s results Wednesday. Sales more than tripled in the company’s fiscal fourth quarter from a year earlier and are projected to do so again in the current period. Earnings surged more than eightfold. The results exceeded analyst expectations. - Nvidia Sales Reach New Heights as Company Forecasts Bigger AI Boom, WSJ
The stock market has become a bit of a popularity contest over the past couple of years with investor sentiment increasingly growing its share of a company’s valuation. The prior darling, Tesla, had dominated daily U.S. stock trading, peaking above $35 billion in turnover several times in recent years. If we compare the stock market to Instagram, we can think of turnover as engagement. The most valuable companies, Microsoft and Apple, are Ronaldo and Messi, the two most followed accounts. But when we think of engagement, it’s the accounts making the most noise that make headlines. When this happens, the question is always if the noise can sustain its growth, and while Elon is good at being loud, Nvidia’s growth may be here to stay:
“Some investors have been scared to buy because they think the stock is too expensive, but that’s been a huge mistake,” said James Demmert, chief investment officer at Main Street Research. “Every time it reports, the P/E shrinks because the E ends up being so much stronger than people expect.”
Put another way, Nvidia’s earnings have been growing even faster than the stock.
Nvidia has seen its valuation slide since the middle of 2023, even amid a record-breaking rally for the shares, because of its massive earnings growth. In the fiscal fourth quarter, the chipmaker reported a whopping 486% year-over-year growth in earnings per share excluding certain items, with the $5.16 figure handily beating analyst estimates of $4.60. Its forecast for first quarter revenue of about $24 billion was also a big beat. — What Bubble? Nvidia Profits Are Rising Even More Than Its Stock, Bloomberg
Zooming out, we know that AI is behind Nvidia’s hype, which is why Goldman’s trading desk called it the “most important stock on Earth.” Post-earnings, we now know that their growth is real and the AI boom will likely continue for some time. It’s not all roses though, as we also know they have a high reliance on Taiwan Semiconductor Manufacturing (TSM) to produce the chips it designs, a risk I previously discussed back in August.
So what’s next? With Nvidia having the first-mover advantage, they’ve essentially cornered the market for AI chips which makes hardware the next frontier.
Hardware
What makes Nvidia’s chips so precious is the flexibility of their GPUs. Quick history lesson, back in the 90s and into the 2000s, Nvidia was only popular amongst gamers because they made the best graphic chips. What changed everything was the launch of its CUDA platform in 2007, which allowed developers to use GPUs for general-purpose computing. This allowed computers to work more efficiently, and because of how complex AI models are, these were the only GPUs that could handle the workload. I’m currently working on a broader essay on AI that will touch on the complexities involved with matrices which are central to AI models. For now, just know that their GPUs ability to excel in parallel processing significantly accelerates these matrix computations. But now that all eyes are on AI, other companies are starting to emerge within the hardware space — say hello to Groq and the LPU:
The Language Processing Unit (LPU) is a new type of end-to-end processing unit system that provides the fastest inference for computationally intensive applications with a sequential component, such as AI language applications (LLMs). The LPU is designed to overcome the two LLM bottlenecks: compute density and memory bandwidth. An LPU has greater compute capacity than a GPU and CPU in regards to LLMs. This reduces the amount of time per word calculated, allowing sequences of text to be generated much faster. Additionally, eliminating external memory bottlenecks enables the LPU Inference Engine to deliver orders of magnitude better performance on LLMs compared to GPUs. - Groq About Us
Their whole thesis lies behind the idea that AI requires new chip architecture. Moreover, with OpenAI CEO Sam Altman trying to raise up to $7 trillion (fk it why not 8) to build AI infrastructure, which would rival Nvidia, the competition is only just starting. This is good old Capitalism at work: competition leads to new growth which means sustained value. So unless shit hits the fan, the AI boom is here to stay.
Thank you
One strategy to be successful in business is to sell the shovels to the miners of the gold rush. During my research, I came across a reddit thread that described Nvidia’s current hype to that of Cisco’s in the 90s; by the end of the dot-com bubble in 2000, Cisco had increased to $500 billion, surpassing Microsoft as the world's most valuable company. At the time, Cisco was essentially the supplier of the internet selling industrial routers and switches to telecom providers while Microsoft was at the cutting edge of of providing the software for users. History may not repeat itself, but it does like to rhyme. With Microsoft’s endless funding of OpenAI, they once again find themselves at the very top, providing software for users against the backbone of Nvidia. As always, if you have any questions, want more explanations, or strongly disagree, comment below, follow me on Twitter (X), follow me on Threads, follow me on TikTok, or shoot me an email.
Disclaimer: These views are my own, and do not necessarily reflect the views of any organization with which I am affiliated with.