Crusoe runs 512 AMD MI355X GPUs at 5.75M tok/s in largest MLPerf inference entry
wkmyrhang · x · 2026-09-17
Crusoe's MLPerf Inference v6.1 submission served gpt-oss-120b and DeepSeek-R1 on 512 AMD Instinct MI355X GPUs — the largest MI355X entry in MLPerf history — hitting 5.75M tokens/s. Throughput scaled near-linearly from 8 to 512 GPUs at >90% of ideal over standard 400Gb Ethernet, with no InfiniBand/RoCE or cross-node RDMA needed, running as a plain Kubernetes job. Manifests are open-sourced for reproduction.
More from Infra
- Jensen Huang says safety-testing data centers will become AI's third compute demand pillar — rohanpaul_ai · 2026-09-17
- Jensen Huang: a 1GW NVIDIA AI factory costs $50-60B but generates ~$50B in annual rental revenue — rohanpaul_ai · 2026-09-17
- Ilya warns neoclouds' weak cybersecurity invites rogue AI agents to hijack compute — Miles_Brundage · 2026-09-17
- AMD's free AI Developer Program: $100 cloud credits, Discord access, hardware raffles — wkmyrhang · 2026-09-17
- After AWS me-central-1 loss, dev jokes about explaining the outage to Codex weekly — andersonbcdefg · 2026-09-17
- The RAM Crisis Is Only Just the Beginning as AI Demand Squeezes Supply — perelin · 2026-09-17