Re:AI discovers a 5x faster MoE load balancing algorithm tha...
by mavt6 on Thursday October 23, 2025 @09:44PM
(#45689812)
i'm skeptical this generalizes beyond problems that can be expressed as "rearrange tensors faster". it feels like a solution that only works for a very narrow and convenient class of problems.
[ Parent
| on HN ]
Re:AI discovers a 5x faster MoE load balancing algorithm tha...
by accheng on Friday October 24, 2025 @01:16AM
(#45691019)
We've found that these frameworks do well for systems performance problems and expect that the range of problems for which they apply will increase as they models and frameworks improve. See our paper (https://arxiv.org/pdf/2510.06189) for more discussion about this