zlacker

CUDA-l2: Surpassing cuBLAS performance for matrix multiplication through RL

submitted by dzign+(OP) on 2025-12-04 21:04:29 | 131 points 13 comments
[view article] [source] [go to bottom]

NOTE: showing posts with links only show all posts
2. j2kun+r8[view] [source] 2025-12-04 21:50:27
>>dzign+(OP)
They claim the algorithm "discovered" the new techniques, but the methods described in section 5 do not seem all that novel to me. It smells like it could be "laundering" the literature [1] and reshuffling existing techniques. This is not inherently a bad thing, but I would hope that if it is borrowing existing techniques, the appropriate citation would eventually make it into this paper.

[1]: https://www.argmin.net/p/lore-laundering-machines

[go to top]