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'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.
Evan Conrad describes the order book as a market for breaking leases. Batch inference can accept interruption in minutes; live inference customers may need months of notice and face SLA penalties. Those interruption rights price differently, similar to power markets' day-ahead and live markets. The terms are the product.
Evan Conrad says SF Compute performs cloud-like work—data-center site visits, bill-of-materials design, cluster architecture, rack rollout, and backplane management of PDUs, switches, DPUs and BMCs—but refuses to own hardware. Ownership would let it manipulate the forwards market, so it stays independent like Switzerland.
Evan Conrad says early H100s and GB200s broke or had firmware bugs, so sophisticated buyers may pay more for units deployed after a certain date to avoid first shipments. That cohort risk becomes another tradable property in SF Compute's order book, alongside chip generation and storage.
Evan Conrad says compute's forward curve can be backwardated or contango and often looks humped. Unsold capacity approaching delivery cannot be sold in the past, so it drops price to clear; later capacity may price lower because buyers pay for shorter lead time. He compares expiring capacity to discounted milk.
Evan Conrad says SF Compute shipped the ability for customers to subdivide their InfiniBand fabric without asking the operator. That lets platform-as-a-service companies redraw cluster boundaries on the fly and scale up or down into the market. He calls it virtualizing the data center and making it programmable.
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.
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.
