https://www.sequoiacap.com/article/ais-600b-question/
What has changed since September 2023?
1.
The supply shortage has subsided Late 2023 was the peak of the GPU supply shortage. Startups were calling VCs, calling anyone that would talk to them, asking for help getting access to GPUs. Today, that concern has been almost entirely eliminated. For most people I speak with, it’s relatively easy to get GPUs now with reasonable lead times.
2 GPU stockpiles are growing:Nvidia reported in Q4 that about half of its data center revenue came from the large cloud providers. Microsoft alone likely represented approximately
22% of Nvidia’s Q4 revenue. Hyperscale CapEx is reaching historic levels. These investments were a major theme of Big Tech Q1 ‘24 earnings, with CEOs effectively telling the market: “We’re going to invest in GPUs whether you like it or not.” Stockpiling hardware is not a new phenomenon, and the catalyst for a reset will be once the stockpiles are large enough that demand decreases.
3.
OpenAI still has the lion’s share of AI revenue: The Information recently reported that OpenAI’s revenue is now
$3.4B, up from $1.6B in late 2023. While we’ve seen a handful of startups scale revenues into the <$100M range, the gap between OpenAI and everyone else continues to loom large
The $125B hole is now a $500B hole: In the last analysis, I generously assumed that each of Google, Microsoft, Apple and Meta will be able to generate $10B annually from new AI-related revenue. I also assumed $5B in new AI revenue for each of Oracle, ByteDance, Alibaba, Tencent, X, and Tesla. Even if this remains true and we add a few more companies to the list, the $125B hole is now going to become a $500B hole.
It’s not over—the B100 is coming:Earlier this year, Nvidia announced their B100 chip, which will have
2.5x better performance for only 25% more cost.
One of the major rebuttals to my last piece was that “GPU CapEx is like building railroads” and eventually the trains will come, as will the destinations—the new agriculture exports, amusement parks, malls, etc. I actually agree with this, but I think it misses a few points:
Lack of pricing power: In the case of physical infrastructure build outs, there is some intrinsic value associated with the infrastructure you are building. If you own the tracks between San Francisco and Los Angeles, you likely have some kind of monopolistic pricing power, because there can only be so many tracks laid between place A and place B. In the case of GPU data centers, there is much less pricing power. GPU computing is increasingly turning into a commodity, metered per hour. Unlike the CPU cloud, which became an oligopoly, new entrants building dedicated AI clouds continue to flood the market. Without a monopoly or oligopoly, high fixed cost + low marginal cost businesses almost always see prices competed down to marginal cost (e.g., airlines).
Investment incineration: Even in the case of railroads—and in the case of many new technologies—speculative investment frenzies often lead to high rates of capital incineration.
The Engines that Moves Markets is one of the best textbooks on technology investing, and the major takeaway—indeed, focused on railroads—is that a lot of people lose a lot of money during speculative technology waves. It’s hard to pick winners, but much easier to pick losers (canals, in the case of railroads).
Depreciation: We know from the history of technology that semiconductors tend to get better and better. Nvidia is going to keep producing better next-generation chips like the B100.
1 Winners vs. losers: I think we need to look carefully at winners and losers—there are always winners during periods of excess infrastructure building. AI is likely to be the next transformative technology wave, and as I mentioned in the last piece, declining prices for GPU computing is actually good for long-term innovation and good for startups. If my forecast comes to bear, it will cause harm primarily to investors.