Get your free personalized podcast brief

We scan new podcasts and send you the top 5 insights daily.

SF Compute's order book has small volume but concentrated liquidity. Evan Conrad tracks whether a customer who spends $100 buying capacity can resell about $25 on average, and measures liquidity by whether it solves the commercial customer's problem rather than by counting makers. He prefers that customer-resale metric to high-minded maker counts.

Related Insights

Firms like OpenAI and Meta claim a compute shortage while also exploring selling compute capacity. This isn't a contradiction but a strategic evolution. They are buying all available supply to secure their own needs and then arbitraging the excess, effectively becoming smaller-scale cloud providers for AI.

Greg Brockman simplifies OpenAI's business to its most fundamental level: buying or building massive amounts of compute and reselling it with an intelligence layer on top. This framing reveals that their primary growth vector and constraint is access to computation, making their core operation a margin-based resale of processing power.

Evan Conrad estimates a B300 can make about $40/hour selling inference at full utilization, or $25/hour at 50%, while B300s rent for $5-6/hour. He says one of those numbers must change, calling it a huge arbitrage. That gap shapes willingness to pay for future capacity.

The head of AI at Hudson River Trading highlights a practical barrier to creating a financial market for compute. For serious training, the minimum "lot size" is thousands of GPUs, not a small, fungible unit. This makes it difficult to standardize a contract and create liquidity, unlike commodities with smaller, interchangeable units.

Kalshi uses market makers to solve the cold-start problem and bootstrap liquidity for new contracts. However, as a market becomes more successful and organic volume grows, the percentage of market maker participation intentionally decreases. Their role is to ignite the flywheel, not to be the engine itself.

Evan Conrad says Nvidia benefits from an independent compute market but can't build it: Nvidia would not be neutral enough for traders, and without trading liquidity lenders won't finance against forwards. Hyperscalers and clouds like CoreWeave also can't, because they have too much demand-side inference and cloud business to give up.

Evan Conrad says two VMs from one operator can differ by 40% in kernel-launch performance, and storage, WAN transit, preemption and chip generation also change value. Rather than wait for standardized contracts, SF Compute lets buyers and sellers attach multiple properties to orders, accepting wider spreads and arbitrage for flexible buyers.

SF Compute argues that for a true compute market to exist, it needs an independent operator that handles physical settlement. This means running the data centers to standardize quality and earn Wall Street's trust, much like an electrical grid operator.

Creating a true commodity market for AI compute is highly challenging because compute is not uniform. Variations in chips, data center design, and specific customer needs make it difficult to establish a standardized physical delivery product or a trusted index, which are prerequisites for a liquid, tradable market.

Evan Conrad argues financializing compute before physical settlement creates a market that is easy to manipulate and carries basis risk: the hedge price may not match actual deliverable capacity. SF Compute co-designed its order book with physical constraints; Conrad says cash settlement is good only if derived from physical settlement.