Qwen@Alibaba_Qwen
31AI 编辑部评分,满分 100
2026-08-15 00:58· 29分钟前
AI 导读

通义千问 Qwen3.8 已通过 TokenSpeed 推理引擎实现大规模运行,延迟低、速度快。LightSeek 作为 Day-0 开源推理引擎合作伙伴,针对该 2.4T 参数模型优化了跨节点 DP/EP 扩展,性能比 TP16 快 30% 以上,并采用 DSpark 投机解码与单 CUDA graph 优化。

Excited to see Qwen3.8 running at scale with TokenSpeed! 🚀 Light on latency, big on speed. Kudos to LightSeek for the fantastic Day-0 support! @lightseekorg

LightSeek FoundationWe're proud to be the Day 0 open-source inference engine partner for @Alibaba_Qwen 3.8. To serve this 2.4T-parameter model across multi-node @NVIDIAAI Blackwell...

来源:Qwen · x.com

Qwen · @Alibaba_Qwen · X·2026-08-15 00:58·29分钟前
AI 导读

通义千问 Qwen3.8 已通过 TokenSpeed 推理引擎实现大规模运行,延迟低、速度快。LightSeek 作为 Day-0 开源推理引擎合作伙伴,针对该 2.4T 参数模型优化了跨节点 DP/EP 扩展,性能比 TP16 快 30% 以上,并采用 DSpark 投机解码与单 CUDA graph 优化。

Excited to see Qwen3.8 running at scale with TokenSpeed! 🚀 Light on latency, big on speed. Kudos to LightSeek for the fantastic Day-0 support! @lightseekorg

LightSeek FoundationWe're proud to be the Day 0 open-source inference engine partner for @Alibaba_Qwen 3.8. To serve this 2.4T-parameter model across multi-node @NVIDIAAI Blackwell...

来源:Qwen· x.com