AMD MI355X hits up to 2x perf-per-dollar vs B200 on Kimi K3 agentic inference
AnushElangovan · x · 2026-11-01
SemiAnalysis reports that AMD's Anush Elangovan and his software team achieved up to 2x better performance per dollar than the B200 on Kimi K3 agentic inference, via tuning AITER kernels, MonoKernel decode, and vLLM work. MI355X's 2304GB HBM per server also lets it avoid the slow 8x400Gbit/s scale-out networking the B200 (under 1.5TB HBM) requires.
More from Infra
- Analyst bets on GPU overbuild, pricing idealized compute at $38/hour — bookwormengr · 2026-11-01
- NVIDIA ships an 81GB FP4 rewrite of DeepSeek-V4.1-Flash with near-identical coding scores — evisoft · 2026-11-01
- Commentary: AI doesn't repeal economics — data center spend must prove its value — DavidLinthicum · 2026-11-01
- Musk confirms Colossus 2 will hit ~1.2M Blackwell GPUs by end of 2025, 5x Colossus 1 — XFreeze · 2026-11-01
- Ex-Nvidia exec: Meta's 40% idle GPUs could flood market, NeoClouds in trouble — SumitGup · 2026-11-01
- CI cost showdown: same 20K minutes — GitHub $120 vs Blacksmith $80 vs Celesto $13 — aniketmaurya · 2026-11-01