AMD MI355X Beats NVIDIA B200 in Kimi K3 Deployment with 952 tok/s
adrianscottcom · x · 2026-08-03
Engineers successfully deployed the Kimi K3 model on a single AMD MI355X server. Leveraging AMD's memory advantage, the single node achieved a total throughput of 952 tok/s and 118 tok/s for a single user.
Tests show this node's total throughput is nearly 4x that of a 16×B200 setup, with better performance per dollar than both B200 and B300. Additionally, AMD's ROCm stack proved mature, working mostly out-of-the-box with only minor fixes needed.
Related event: AMD MI355X Runs Kimi K3 at 952 tok/s, Outperforming B200(6 posts)→
More from Infra
- China's AI boom drives up to 660% profit surges for domestic chip firms; GPU maker MetaX turns profitable — pstAsiatech · 2026-10-16
- Qdrant tests Perplexity's visual PDF embeddings: 21x smaller vectors, 99.9% after rescoring — jrxna · 2026-10-16
- When Does an Agent Actually Need an API Gateway? Reddit Weighs Direct Calls vs. Unified Endpoints — TurbulentStiffness · 2026-10-16
- Taiwan to build $4B AI manufacturing complex in Texas to back Nvidia and TSMC — Polymarket · 2026-10-16
- When the KV Cache Is Slower Than the GPU: Remote KV Tiers Backfire in Production — MultiverseComputingCAI · 2026-10-16
- GitHub rewrote the 800k-line Copilot runtime in Rust using AI agents, cutting CPU and memory by over 90% — bibryam · 2026-10-16