In the news
Muse Spark Open Ultrafast: Pushing beyond 1,000 Tokens per second
LMSYS · Published · 3 min read
In 30 seconds
- What happened
- LMSYS achieved 1,109 tokens per second on Muse Spark Open using megakernels, doubling the baseline 517 tok/s through fused GPU operations.
- Why it matters
- Matters for engineers optimizing LLM inference on modern GPUs, especially those targeting high-throughput serving with tensor parallelism.
- Watch out
- Megakernels require 8000+ lines of low-level CUDA and PTX code; instruction cache conflicts and register pressure create non-obvious performance cliffs.
- token
The patterns behind this
- Agent-Readable Web (llms.txt / NLWeb)
- Generative UI (Agent-Rendered Interfaces)
- HTTP-Native Micropayments (x402)
Each one covers how the technique works, when it earns its cost, and where it breaks.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.