# Prime Intellect makes DeepSeek's Mega MoE kernel trainable

> Original title: Introducing Prime Mega MoE.

- Source: Prime Intellect
- Published: 2026-10-29T17:41:34.000Z
- Bad Influence: https://badinfluence.news/items/kuhiymwmrghkei7cl3b26cjht
- Original: https://x.com/PrimeIntellect/status/2115861613261815883

## Summary

Prime Intellect says it made DeepSeek's Mega MoE kernel trainable. On two Qwen3 models, it reports up to 55% faster MoE training than TorchTitan and up to 34% faster hosted training with prime-rl.

## Post

Introducing Prime Mega MoE.

We made DeepSeek's Mega MoE kernel trainable. The results on two Qwen3 models:

• Up to 55% faster MoE training than TorchTitan
• Up to 34% faster hosted training with prime-rl

[Video](https://badinfluence.news/api/img-proxy?u=https%3A%2F%2Fpbs.twimg.com%2Famplify_video_thumb%2F2115861599693242368%2Fimg%2Fxp42Be__veut1imf.jpg&mode=thumb&exp=1793491200&sig=8a0b9c13a1d4b89f)