Why AI Compute Comes in Rectangles — Erik Bernhardsson, Modal CEO
- Who
- Erik Bernhardsson
- What
- explains the infrastructure and economics behind Modal’s growth and its approach to elastic GPU capacity
- Quote
In this conversation, Erik joins Vibhu to explain the infrastructure and economics behind that growth—and why demand for compute rarely looks like the contracts used to buy it.
What happened
Modal CEO Erik Bernhardsson explains why fixed GPU contracts poorly match fluctuating AI workloads and how Modal pools customer demand to offer elastic compute. Modal reports crossing $1 billion in annualized run rate and raising a $750 million Series D at a $16B valuation. Bernhardsson says sandboxes now generate roughly a third of revenue, while faster startup and snapshotting could support shorter, more efficient execution cycles for agents.
Read next
See the full storyModal CEO Explains Elastic Compute Strategy as Company Reports $1 Billion Annualized Run RateThe story so far
Modal reports crossing $1 billion in annualized run rate and raising a $750 million Series D at a $16 billion valuation, while CEO Erik Bernhardsson describes pooling customer demand as the basis for its elastic compute offering. Bernhardsson says fixed GPU contracts poorly match fluctuating AI workloads. Modal’s approach pools demand across customers to offer capacity that can adjust to their needs. He says sandboxes now generate roughly a third of Modal’s revenue, and faster startup and snapshotting could enable shorter, more efficient execution cycles for agents.
Keep up with AI, one email a day
Model releases, price changes and compute developments, with original sources. Free, confirmed by email, and easy to unsubscribe. Read a past issue · 2026-10-22