Designing Compute Markets | Kavish Garg
kavishgarg.com · 3,658 words · saved by 3 readers
A proposal for workload-native compute markets, forward auctions, spot pooling, and capacity markets.
We have been increasingly crunched for compute capacity over the past few months. Everyone, from early-stage startups to established hyperscalars, shares this sentiment . Market pricing does indeed reflect this narrative: the prices for 1-year H100 leases have risen 35% over the past four months. Amidst this capacity crunch, we have spontaneously seen market-making or brokering mechanisms emerge. From VC firms running compute pools to independent compute brokers slinging H100s, everyone has seemingly made “compute” an agenda. Many cloud aggregators are also working on compute allocation in a m
saved by
related reading
- Could Compute Become a Commodity? Lessons from Oil, Power, and Spectrumformvfunction.substack.com
- $2 H100s: How the GPU Bubble Burst - by Eugene Cheahlatent.space
- The birth of the AI compute marketdavefriedman.substack.com
- Clouded Judgement 7.3.26 - The End of Compute Scarcity? Not So Fastcloudedjudgement.substack.com
- On Making a Trading Market for "Compute"ai-street.co
- Why compute might get 10x more expensive in coming yearsdwarkesh.com
- New Article: Wiring Capital to Computeletter.palladiummag.com
- Ornn | Financial Products for Computeornn.trade
- Dave Friedman on X: "The compute market is interesting. The AI labs and hyperscalers can't get enough compute. They're all buying/leasing GPUs and using crazy financial structures/circular deals, etc. And Don Wilson and similar people are ax.com
- The hidden risk of AI computedavefriedman.substack.com
- The Most Expensive GPU is The One You Didn't Hedge - ADM Investor Servicesadmis.com
- The Most Expensive GPU is The One You Didn't Hedge - ADM Investor Servicesadmis.com