How CoreWeave Sees the Market for Compute Right Now

June 8, 2026

AI Summary

5 min read

How CoreWeave Sees the Market for Compute Right Now

The AI inference market is moving so fast that companies are burning through their entire annual AI budgets in months, not years. Uber reportedly exhausted its 2026 AI budget in four months, and one unnamed client spent half a billion dollars in a single month after failing to set usage limits. These headlines capture a moment where the AI compute market has shifted from experimental dabbling to a full-blown corporate spending spree—and the CFOs are starting to notice.

The Demand Picture: Unrelenting but Shifting

Brandon McBee, CoreWeave's co-founder and chief development officer, sees no pullback in demand. The company supports nine of the top ten AI labs globally (excluding China), and its financial services client backlog alone is approaching $10 billion. The customer base has diversified enormously from the early days when CoreWeave had just a handful of large clients. In Q4 alone, the company added twice as many enterprise logos as in any previous quarter.

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

Never miss an episode of Odd Lots

Get every new episode summarized in your inbox — free, ~5 minutes to read.

No spam. Unsubscribe anytime.

What you'll learn

1 (01:49) **Episode Introduction & The "Inference Update" Thesis** - Hosts Joe Weisenthal and Tracy Alloway frame the episode as a necessary pulse-check on the rapidly evolving AI inference market, noting that corporate CFOs are getting sticker shock from compute budgets.
2 (06:21) **Guest Introduction: Brandon McBee of CoreWeave** - Co-founder and Chief Development Officer of CoreWeave, a leading "neo-cloud" provider, returns to discuss the state of AI compute demand.
3 (07:25) **Is Inference Demand Slowing?** - Brandon sees no pullback; the "authentic and foundational demand" is unrelenting, driven by a product breakthrough in early 2025.
4 (09:52) **Customer Mix Evolution** - The customer base has diversified enormously from a handful of hyperscalers and AI labs to three distinct buckets: hyperscalers, AI labs (9 of top 10), and a rapidly growing enterprise base.
5 (12:49) **Direct Enterprise vs. AI Lab Customers** - Clarifies that financial services clients like Jane Street come directly to CoreWeave for their own proprietary models, not through an AI lab.
6 (17:27) **Training vs. Inference & Lengthening Contracts** - AI labs are signing longer (5-year) take-or-pay contracts for specific GPU generations (Hopper, Blackwell), indicating sustained demand for frontier model training at scale.
7 (20:12) **Vera Rubin & GPU Generational Shifts** - Explains NVIDIA's next architecture (Vera Rubin) and the increasing complexity of deployments (e.g., liquid-cooled 72-GPU racks for Blackwell).

+ Full timestamped outline available in the app

Show Notes

When we last spoke to Brannin McBee, the co-founder and chief development officer of cloud company CoreWeave, his business was not yet public and sourcing GPUs was a key constraint on growth. But three years later, things look pretty different. CoreWeave IPOed and has been raising money in the bond market too, as well as signing more deals with chipmaker Nvidia. In fact, investors have basically been throwing money at all-things-AI. But there are persistent bottlenecks to further growth. Chip supply is still scarce, but so are transformers and electricity. In this episode, we catch up with Brannin on everything he's seeing in the market for compute right now, including leases, Nvidia's new Vera Rubin systems, demand for training versus inference, and the possibility of standardizing the market for compute.

Only http://Bloomberg.com subscribers can get the Odd Lots newsletter in their inbox each week, plus unlimited access to the site and app. Subscribe at bloomberg.com/subscriptions/oddlots

Subscribe to the Odd Lots Newsletter