Qwen3 hits 3.3k tokens/s on single Nvidia GH200 with 64 concurrent requests
MaziyarPanahi · x · 2026-08-15
Benchmarks reveal impressive performance for Qwen3-8B-27B-FP8 on a single Nvidia GH200:
- Throughput: 3,300 tokens/s with 64 parallel requests.
- Latency: First streamed events landed within 10ms.
- Scale: Handled 10 concurrent streaming requests with max 16K output and 262K context.
- Stability: All requests completed successfully.
Tests were conducted on vLLM via Lambda API.
Related event: Single GH200 runs Qwen 27B FP8: 10 concurrent streams, 10ms first token(5 posts)→
More from Infra
- TokenSpeed MLA 2 hits Vera Rubin: up to 3.44× faster MLA decode for Kimi K3 — zhyncs42 · 2026-10-16
- US vs Chinese hyperscalers: Americans fund and schmooze, Chinese demand your BoM and cap you at 7% margin — jwt0625 · 2026-10-16
- How LiteLLM routes one OpenAI-shaped API across every model provider — Mahmoud_Zalt · 2026-10-16
- OpenAI-Broadcom custom AI chip Jalapeño named a TIME Best Invention of 2026 — itsclivetime · 2026-10-16
- Developer live-runs continual pretraining of a Luganda LLM on Colab, 41% done with loss dropping fast — weskambale · 2026-10-16
- Jensen Huang: Nvidia charges ~$thousands per GPU yearly for software, a major revenue line — rohanpaul_ai · 2026-10-16