Cursor 已开源 Mixture-of-Kittens (MoK)，这是一个专为 NVL72s 设计的 MoE 训练超级内核。该内核融合了专家混合模型的通信和计算，与现有基线相比，性能最高可提升 2.37 倍。MoK 目前支持 Cursor 上数万个 GPU 的模型训练，将端到端训练吞吐量提高了 1.41 倍。 AI
影响 此次开源旨在降低人工智能研究的门槛，使更多实验室能够更高效地训练模型。
排序理由 AI IDE Cursor 开源了训练超级内核。
AI 生成摘要 · Google Gemini · 来自 3 个来源。 我们如何撰写摘要 →
完整方法见我们的编辑标准。
Our hope is that this lowers the barrier to AI research, so more labs can train models efficiently. Here's how we built it: https://t.co/EaC2z7Mw3a
MoK now powers training across tens of thousands of GPUs at Cursor. In production, it raised end-to-end training throughput by 1.41x over our previous DeepEP-based stack.
We're open-sourcing Mixture-of-Kittens (MoK), our MoE training megakernel for NVL72s. It fuses all Mixture-of-Experts communication and computation into a single, fully deterministic kernel, and runs up to 2.37x faster than the strongest public baselines. https://t.co/yHu5E6RXp9