AMD 8×MI355X 上 V4.1 Flash 单用户超 2100 tok/s

AI at AMD @AIatAMD原推发布于 2026.10.28

在 X 上打开原推

2,100+ tokens/sec per user on DeepSeek V4.1 Flash, at full model quality, on 8× AMD Instinct MI355X. 🚀 Congrats to @lithos_ai on pushing the frontier of inference with persistent megakernels on AMD. Open hardware + instruction-level control = room to innovate.

这条是引用着另一条写的：「Lithos 引擎在 8 张 MI355X 上跑 DeepSeek V4.1 Flash：单用户 2100+ tok/s」