PulseAugur
中
实时 18:48:57
English(EN) One-Sided Expert-Parallel All-to-All for Mixture-of-Experts Inference over NVLink

新的 NVLinkOneSided 协议加速 GPU 上的 MoE 推理

研究人员开发了 NVLinkOneSided,这是一种新颖的单边专家并行全互连通信协议,专为单个 NVLink 域内的 GPU 上的专家混合 (MoE) 推理而设计。该方法允许 GPU 将 token 直接推送到目标缓冲区,绕过元数据交换并提高效率,尤其是在小批量大小的情况下。在 Blackwell GPU 上进行基准测试,与 DeepEP 相比,NVLinkOneSided 在调度和组合操作方面速度提高了 1.2-3.0 倍,MoE 层速度提高了 2.4 倍,使其在解码任务上甚至可以与 MegaMoE 等融合后端竞争。 AI

影响 加速专家混合模型的推理,可能实现更快、更高效的大型 AI 模型的部署。

排序理由 描述 AI 推理新技术的论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 Hugging Face Daily Papers 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。 我们如何撰写摘要 →

新的 NVLinkOneSided 协议加速 GPU 上的 MoE 推理

本文如何被排名

Signal score
0 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
描述 AI 推理新技术的论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
infra, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
12 days old
Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

报道来源 [1]

  1. Hugging Face Daily Papers TIER_1 English(EN) ·

    面向NVLink上混合专家模型推理的单边专家并行全互联

    Expert parallelism (EP) shards the experts of a Mixture-of-Experts (MoE) layer across GPUs, so every MoE layer requires two all-to-all communications, namely dispatch and combine. Compared with training, inference processes relatively few tokens per GPU at a time, and at small ba…