The 'compute spread': H100 rental vs open-source LLM token prices show -0.46 correlation
AccBalanced · x · 2026-10-24
Borrowing the commodities concept of the 'crack spread' (oil vs refined product prices), Antonio García Martínez argues AI has an analogous 'compute spread' between hardware compute costs and inference token prices. Data shows H100 rental vs open-LLM token daily changes correlate at roughly -0.46 over 29 days — noisy but correlated, expected to stabilize as markets mature. Trillion-dollar empires and nation-states, he says, will rise and fall on this spread.
More from Infra
- Flappy founders on CNBC: co-optimizing the AI stack unlocks far more chip value — DavidCahn6 · 2026-10-24
- Fireworks ships 7 new open models in two weeks, betting on specialized intelligence — sophiamyang · 2026-10-24
- Every Mac on macOS 27 ships a free offline LLM in Terminal via the fm CLI — dr_cintas · 2026-10-24
- NVIDIA and UW Protein Design team up to release RFD4-Proteina for designed biology — AllThingsApx · 2026-10-24
- Grok does the math: running Kimi K3.1 Pro "locally" takes 18-32 H100s and $0.5-2M in hardware — giffmana · 2026-10-24
- A ~€1600 local LLM rig: i7-14700KF + RX 9070XT hits 70-90 tokens/sec — Snoo-2768 · 2026-10-24