Estimating Compute Revenue and GPU Scale
teortaxesTex · x · 2026-07-15
Based on the "bundle cost," the author estimates that peak load requires about 1000–2000 seconds of GPU time.
Calculated against this metric, the revenue per GPU is roughly $28,000 to $55,700 annually, corresponding to approximately 9,000–18,000 GPUs, or 1–2 SuperPODs. The author adds that if peak loads are more extreme and the 2K estimate is too low, it might require up to 40,000 seconds of GPU time, suggesting the actual scale could be even larger.
Related event: DeepSeek inference economics: sizing the GPU fleet behind its ARR(8 posts)→
More from Infra
- PyTorch details 2.12x faster on-device streaming ASR on Arm SME2 via ExecuTorch — PyTorch · 2026-10-27
- Revenue per GPU can rise as token prices fall, argues Gavin Baker — GavinSBaker · 2026-10-27
- Anthropic S-1: only $19B total compute spend since inception, overbuild warning — bookwormengr · 2026-10-27
- CoreWeave, You.com and NVIDIA lift BrowseComp 22.9% in 3-week RL post-training sprint — RichardSocher · 2026-10-27
- Grid queues make power architecture an AI-factory decision; Exowatt and Redwood offer fixes — ingliguori · 2026-10-27
- Two CVSS 9.9 Wasmtime flaws allow sandbox escape, one ARM64-only since 2020 — basedjensen · 2026-10-27