Despite the massive investments in AI infrastructure by tech giants, AI revenue growth has yet to materialize, indicating a significant gap in the ecosystem’s value to the end user. In fact, David Cahn, an analyst at Sequoia Capital, estimates that AI companies will need to earn about $600 billion annually to fund their AI infrastructure, such as data centers.
Last year, Nvidia generated $47.5 billion in revenue from data center hardware (most of which was compute GPUs for AI and HPC applications). Companies like AWS, Google, Meta, Microsoft, and many others have invested heavily in their AI infrastructure in 2023 for applications like OpenAI’s ChatGPT. But will they recoup that investment? David Cahn thinks this could mean we’re seeing the growth of a financial bubble.
Simple Math
Cahn’s math is relatively simple. First, he doubles Nvidia’s revenue projections to cover the total costs of the AI data center (GPUs account for half; the rest is power, buildings, and backup generators). Then he doubles that amount again to account for a 50% gross margin for end users, such as startups or enterprises buying AI computing resources from companies like AWS or Microsoft Azure, who also need to make money.
Cloud providers, including Microsoft, are investing heavily in GPU inventory. Nvidia said half of its data center revenue comes from large cloud providers, with Microsoft alone contributing about 22% of Nvidia’s revenue in the fourth quarter of fiscal 2024. At the same time, the company sold about $19 billion worth of data center GPUs in the first quarter of fiscal 2025.
The introduction of Nvidia’s B100/B200 processors, which promise 2.5 times better performance at only 25% higher cost, will likely drive new investments and create a new supply shortage.
According to the analyst, OpenAI, which uses Microsoft’s Azure infrastructure, has seen a substantial increase in revenue, from $1.6 billion at the end of 2023 to $3.4 billion in 2024. This growth underscores OpenAI’s dominant position in the market, far ahead of other startups that are still struggling to reach the $100 million revenue mark. Yet investment in AI hardware is on the rise.
Even the big tech companies’ optimistic projections for AI revenue fall short, Cahn says. Assuming that Google, Microsoft, Apple, and Meta each generate $10 billion a year from AI, and that other companies like Oracle, ByteDance, Alibaba, Tencent, X, and Tesla each generate $5 billion, that leaves a $500 billion gap.
The AI industry must learn to win
Investments in AI infrastructure face major challenges. Unlike physical infrastructure, AI GPU computing could become commonplace as new players enter the scene (AMD, Intel, not to mention custom processors from Google, Meta, and Microsoft), especially in inference, leading to intense price competition. Speculative investments often lead to high losses, and new processors quickly devalue old ones, unlike the more stable value of physical infrastructure.
Ultimately, while AI has transformative potential and companies like Nvidia play a crucial role, the road ahead will be long and bumpy, as companies and startups have yet to invent applications that generate money.
Cahn believes the industry needs to temper its expectations of quick profits from AI advances, recognizing the speculative nature of current investments and the need for sustainable innovation and value creation. If it doesn’t, the multi-trillion dollar bubble risks bursting, leading to a global economic crisis, but we’re just speculating here.
