Re:AI Is Writing Its Own Kernels, and They Are 17x Faster
by greeravoctado on Thursday November 20, 2025 @07:29PM
(#45999829)
Usually the rate of overall improvement for this type of optimization is less than Moore law rate of improvement, thus not worth the company investment. 17x micro-benchmarks don't count. Real improvements come from architectural changes, for example: MoE, speculative multi-token prediction, etc.
The expense calculation might be
expense of improvement = (time taken per optimization step * cost of unit time ) / ( speedup - 1)
The expensive heuristic function is saving wall time well also being cheaper in cost of unit time. And as the paper shows the speed up provided for each unit time multiplied by unit cost of time is large.