News for hackers. Stuff that matters.
Re:Google Cloud TPU Multislice Training by aschleck on Friday November 10, 2023 @03:59PM (#38224446)
1979 16 bit flops on an H100 is with sparsity. See footnote 2 on https://www.nvidia.com/en-us/data-center/h100/. You should be halving it for non-sparse flops.
Re:Google Cloud TPU Multislice Training by YetAnotherNick on Saturday November 11, 2023 @03:11AM (#38228450)
GP is correct. With sparsity it is 3958. 1979 Tflop/s is without sparsity.
Re:Google Cloud TPU Multislice Training by emu on Saturday November 11, 2023 @04:45AM (#38228835)
No, it is not. That's the sparse fp8 flop number, but you need to ignore sparsity and compare bf16 flops not fp8 flops for the comparison the ancestor post is making.