inference infrastructure
2 articles
SGLang 0.5.20 Upgrade Guide: CUDA, APIs and Rollback
A 2026 SGLang 0.5.20 upgrade guide covering CUDA 12 artifact retirement, Responses API storage, removed controls, platform compatibility, canary acceptance tests and a validated rollback plan.
October 17, 2026• 23 min read
sglang 0.5.20sglang upgradecuda 13
MLPerf Inference 6.1: A GPU Procurement Guide
A 2026 guide to reading MLPerf Inference 6.1 for GPU procurement, with cohort filters, exact result-row examples, scale-efficiency math, and a reproduction pilot worksheet.
September 18, 2026• 20 min read
mlperf inference 6.1gpu procurementmlperf benchmarks