TLDR

Lightseek says its TokenSpeed-powered Kimi K3 setup reached 216 output tokens per second per user on eight AMD Instinct MI355X GPUs. It claims that is 1.4 times ATOM's decode speed at C1 while running at higher precision. Lightseek says it measured the result on a multi-turn SWE-smith agentic workload with 50K-plus context and 10–15 turns.

Announcement

Kimi K3 reportedly reached 216 output tokens per second per user on eight AMD MI355X GPUs

Lightseek says the TokenSpeed-powered run delivered 1.4 times ATOM's decode speed at C1 while running at higher precision.

2 Sources, 2h ago, first seen 2h ago

TLDR

Today's Rank

—

Not ranked yet

Combined views

—

2 Sources, first seen 2h ago

— likes— comments— saves— reposts

Sentiment

Positive——Negative

Summary

Not enough discussion yet.

No sentiment analysis available yet.

Featured Source