FrontierAugust 4, 2026via MarkTechPost

Cursor Open-Sources Mixture-of-Kittens (MoK): A Deterministic MoE Training Megakernel for GB300 NVL72 Racks

Why it matters

A significant optimization in frontier model training (MoE efficiency gains), but with a harsh availability gate: the kernel requires Blackwell-class GPUs that most practitioners cannot access. Relevant to labs training large models and to the broader frontier-lab race on training efficiency; less immediately actionable for most practitioners.

Key signals

  • Cursor Research open-sourced Mixture-of-Kittens (MoK)
  • MoE training megakernel achieves 2.37x speedup vs. public baseline
  • Fuses MoE communication and computation into single deterministic kernel
  • Requires Blackwell SM100 or SM103 GPUs (NVL72 racks only)
  • Powers Cursor's Composer models
  • Published August 4, 2026

The hook

Cursor open-sources MoK, a 2.37x faster MoE training kernel—but only for teams with NVL72 racks.

Cursor Research has open-sourced Mixture-of-Kittens (MoK), the MoE training megakernel behind its Composer models. MoK fuses all mixture-of-experts communication and computation into a single deterministic kernel, and runs up to 2.37x faster than the strongest public baseline on GB300 NVL72 racks. I

The week's key stories, every Friday.

For practitioners and enthusiasts — free, in your inbox.

Free forever. No spam.

Cursor Open-Sources Mixture-of-Kittens (MoK): A Deterministic MoE Training Megakernel for GB300 NVL72 Racks | KeyNews.AI