Morning Brief · November 2nd 2026
AMD's MI355X serves Kimi K3 for up to 32% less than Nvidia's GB200, SemiAnalysis's tests show
On SemiAnalysis's InferenceX , serving Moonshot's 2.8-trillion-parameter Kimi K3 on replayed coding-agent sessions cost $0.154 per million on AMD's MI355X against $0.225 on Nvidia's GB200 at 165 tokens per second per user, and $0.050 against $0.068 at 71. The gap is narrowest at 118 tokens per second, $0.138 against $0.165. The costs assume large-hyperscaler ownership at $1.50 an hour per MI355X and $1.86 per GB200.
The percentages printed on the page are larger than its own dollar figures support; we computed 16% to 32% from the dollars. AMD reports on Tuesday. SourcesA
Sources
- AInferenceX: Kimi K3 on GB200 NVL72 vs MI355X, cost per dollar (SemiAnalysis)inferencex.semianalysis.com
This story ran in the Morning Brief of November 2nd 2026.