TLDRocket
Sign in

Cursor Research Open-Sources Mixture-of-Kittens Deterministic MoE Training Megakernel

Open source release Provisional 92% confidence first seen

Cursor Research released Mixture-of-Kittens (MoK), an open-source mixture-of-experts training kernel that consolidates MoE communication and computation into a single deterministic megakernel optimized for NVIDIA Blackwell GPUs. The kernel achieves up to 2.37x higher throughput than existing baselines and is available under Apache-2.0 license for organizations with large-scale GPU infrastructure.

Cursor Cursor Research Mixture-of-Kittens DeepSeek-V3 Meta NVIDIA PyTorch Rubin TensorRT-LLM

Source coverage

Cursor Open-Sources Mixture-of-Kittens (MoK): A Deterministic MoE Training Megakernel for GB300 NVL72 Racks MarkTechPost [AINews] Megakernels are so dead and so back Latent Space newsletter

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.