Fp8 runs ~100 tflops faster when the kernel name has "cutlass" in it

Re:Fp8 runs ~100 tflops faster when the kernel name has "cut... by mort96 on Friday October 03, 2025 @05:09AM (#45460753)

Are there any archives of that techreport article with images intact?

[ Parent | on HN ]

Re:Fp8 runs ~100 tflops faster when the kernel name has "cut... by haunter on Friday October 03, 2025 @06:28AM (#45461241)

Ah yes they changed the site and URL system after some years, here is the OG one with screenshots

[ Parent | on HN ]