The Programmable Data Center — Evan Conrad, SF Compute
October 20, 2026
AI Summary
5 min readThe Programmable Data Center — Evan Conrad, SF Compute
When Evan Conrad started what would become SF Compute, he was just trying to sublease a GPU cluster — like renting out a hotel room. Then customers asked for cleaning, then room service, then security. "At a certain point you're just running the whole hotel," Conrad says. That experience taught him something the industry is still learning: compute is not a commodity you can trade like a ticker symbol. It is a physical thing with physical constraints, and any market that pretends otherwise is building on sand.
Physical Settlement: Why Compute Is More Like Oil Than Like a Stock
SF Compute borrows the term "physical settlement" from futures markets — in oil or soybeans, physical settlement means you actually get the oil or soybeans delivered. In compute, it means you get the actual GPU capacity, with all its real-world quirks. Conrad argues this is the only honest foundation for a compute market.
The analogy to oil is instructive. During COVID, oil prices went negative because sellers would pay buyers to take oil off their hands — it was sitting in ports and storage was full. Compute has similar weirdness. A cluster's boiler (so to speak) takes time to cool off and restart. An order book that ignores this will produce prices that have nothing to do with what anyone can actually deliver.
Continue reading the full summary in the app — free to try.
Read Full Summary →Free • No credit card required
Never miss an episode of Latent Space: The AI Engineer Podcast
Get every new episode summarized in your inbox — free, ~5 minutes to read.
No spam. Unsubscribe anytime.
What you'll learn
- 1 (00:00) **The NVIDIA Paradox** - Why NVIDIA can't build SF Compute itself, even though it benefits from the market existing
- 2 (01:21) **From Subleasing to Running the Whole Cloud** - The origin story of SF Compute's evolution from a simple sublease model to full physical settlement
- 3 (03:49) **Physical Settlement: The Core Concept** - Borrowing a term from futures markets to explain how compute delivery actually works
- 4 (05:46) **Market Maturity: Not Very Liquid Yet** - Current state of compute market depth and liquidity metrics
- 5 (07:17) **Why Standardization Is Still Far Away** - The challenge of creating fungible compute contracts when the underlying hardware is wildly heterogeneous
- 6 (09:42) **The Dimensions That Matter** - The key attributes that differentiate compute capacity in the market
- 7 (14:44) **Forward Curves: Three to Five Years Out** - How far out compute markets extend and the shape of the curve
+ Full timestamped outline available in the app
Guests on this episode
Show Notes
This was recorded on the sidelines of the new Compute Trading track at AIE NYC. Catch us next at AIE CODE, our last flagship for the top agentic engineers in the world.
In the early days of cloud, former guest Martin Casado made something so disruptive that Cisco was threatened enough to pay $600m for it, and VMware outbid for $1.3B:
That concept was software-defined networking, which virtualized the routing network by separating the control plane from the data plane. That revolution is now coming for GPUs - where previously running your own supercompute cluster required $50-100m upfront to get commitments and do the entire stack from hardware to firmware to kernels, SF Compute is now enabling you to not only buy or sell GPU hours now and in the future, but also to programmatically control everything from your network to memory.
We last spoke with Evan about commoditizing compute. At AI Engineer New York, we went deeper: what does it take to physically deliver a GPU contract, and what has to be true before a market in compute can become liquid?
From selling a room to running the hotel
Evan describes SF Compute’s expansion with a hotel analogy. A customer wants a room, then asks for cleaning, room service, and security. Eventually you discover that delivering the room means operating the hotel. SF Compute began by trying to sublease a cluster; the company has since built much of the operating capacity of a cloud to deliver the compute traded on its market.
That work is physical settlement: the buyer gets usable compute, rather than simply a cash payment against an index. It brings the physical constraints of a cluster into the market itself. Can a workload be interrupted? How quickly can it recover from storage? Which GPU generation, networking and software configuration does it need? Those details affect both the contract and the price.
Making compute tradeable
Rather than waiting for every GPU cluster to become interchangeable, SF Compute matches orders with different constraints. Evan argues that this makes trading possible while the technology changes, at the cost of spreading liquidity across more variants. We also discuss spot and forward pricing, batch versus live inference, and why an index can become disconnected from the cost of actually buying compute.
Independence is central to Evan’s case for a market operator. He argues that companies with large hardware or inference businesses have incentives that make it harder to run a neutral exchange. We also get into the unglamoro