AMD MI355x beats Nvidia B300 on tokens-per-dollar TCO in AgentX, SemiAnalysis says
AnushElangovan · x · 2026-09-04
Per SemiAnalysis, a new AMD MI355x submission beats Nvidia's B300 on total tokens per dollar (TCO) at lower interactivity ranges on the AgentX benchmark, with credit to engineers from vLLM, AMD, and LMCache. For agent workloads, AMD now holds a local price-performance edge over Nvidia in inference.
More from Infra
- Open-source voice pipeline adds Smart Turn end-of-turn gate before LLM calls — ivan_digital · 2026-09-04
- vLLM team launches Inferact, powers new HUMAIN-M3 Arabic frontier model's inference — woosuk_k · 2026-09-04
- Pedro Domingos: Double LLM efficiency and you should be worth $100B, given Nvidia's math — pmddomingos · 2026-09-04
- Liquid AI launches Nanos: task-specific 350M-2.6B models that run on-device — JosephJacks_ · 2026-09-04
- Miles Brundage: AI outage cascade likely caused by enterprises mass-switching to new models — Miles_Brundage · 2026-09-04
- OpenAI's Astra can now layout and route PCBs, sparking hardware engineering debate — MikePFrank · 2026-09-04