Vibeleaderboard
← All Intel
Intel / post

SemiAnalysis: AMD beats Nvidia on margins for Kimi K3 inference

Source
x.com
Date
SemiAnalysis@SemiAnalysis_

AMD ALERT🚨🚨: On Kimi K3 2.8T, AMD has even better profit margins per gigawatt compared to B300, GB200 NVL72, and B200. This is due to AMD's lower TCO and better performance, leading to better profit margins of 53.6% on Mi355X compared to 44.3% on GB200 NVL72. This is from our open-source Agentic Inference benchmark, AgentX. Shoutout to @roaner & @EmadBarsoumPi 10x engineers for optimizing this 2.8T model!

Context

SemiAnalysis reports that on its open-source Agentic , AgentX, running Moonshot AI's 2.8-trillion-parameter Kimi K3 model, AMD's MI355X GPU produced a 53.6% profit margin per gigawatt of deployed compute, ahead of Nvidia's GB200 NVL72 at 44.3%, and also ahead of Nvidia's B300 and B200 chips on the same measure. SemiAnalysis attributes the gap to AMD's lower total cost of ownership and better measured performance on this specific inference workload. This is a single benchmark result on one model and one workload, not a general claim that AMD outperforms Nvidia across all inference tasks, and it is a data point rather than investment guidance.

Terms in this piece · Glossary
  • inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
  • benchmark — A standard public test set for comparing AI models — the shared scoreboards behind every "model X beats model Y" claim.
More from SemiAnalysis
Recommended reads
Comments

Checking sign-in…

Loading comments…