Earnings watchers eye AI capex
Add Axios as your preferred source to
see more of our stories on Google.

Illustration: Annelise Capossela/Axios
Big Tech earnings for the latest quarter kick off today with Microsoft, and all eyes are on AI-related capex.
Why it matters: What these big companies spend on AI chips has big implications for AI startups and their investors.
- Big cloud providers like Microsoft's Azure and Amazon's AWS are buying up (and making) AI chips not only for their own uses, but also to sell access to those compute resources to other companies, from AI giants like OpenAI to tiny startups.
- VCs poured $38.6 billion into AI and machine learning startups in the U.S. in the first half of 2024, per PitchBook. In all of 2023, they invested $64.6 billion.
The big picture: Nvidia's data center revenue — a proxy for how much the market is spending on AI capex — has been growing at a fast clip.
- Last quarter, it raked in $22.6 billion in data center revenue, up 23% from the previous quarter and up 427% from a year ago (yes, it more than quadrupled since early 2023).
- The company is scheduled to announced its next quarterly results on Aug. 28. Rivals AMD and Arm are scheduled for today and tomorrow, respectively.
Zoom in: Microsoft spent $14 billion on capex last quarter, up 79% and expected to say this afternoon that kept growing.
- Alphabet spent about $12 billion last quarter, up 90% year over year.
- Meta spent $6.7 billion, a bit less than it spent a year prior, though the company raised its capex guidance for the year to $35 billion-$40 billion, up from $30 billion-$37 billion.
Between the lines: There's already some skepticism about this level of AI infrastructure investment and whether we'll ever get the necessary levels of revenue to justify it.
- In a recent blog post, Sequoia Capital partner notes that OpenAI is still generating the bulk of AI-related revenue right now, and its annualized revenue has been pegged at a mere $3.4 billion. Even if major tech companies generate $5 billion-$10 billion in annual AI revenue, it'll barely cover the 2024 infrastructure investment, per his estimate.
- Meanwhile, Barclays estimates that AI capex by 2026 will be sufficient to support 12,000 AI products of the scale of ChatGPT — and it's unlikely we'll get or want 12,000 such apps.
The bottom line: Whatever the numbers we see this week from the Big Tech companies, some venture capital will be incinerated as an indirect result of it.
