gpu-dissection-toolbox/bookmarks.txt at main · plotfi/gpu-dissection-toolbox · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
https://chipsandcheese.com/2023/07/02/nvidias-h100-funny-l2-and-tons-of-bandwidth/

https://developer.nvidia.com/blog/maximizing-unified-memory-performance-cuda/

https://developer.nvidia.com/blog/nvidia-hopper-architecture-in-depth/

https://stackoverflow.com/questions/36416843/how-to-write-a-pointer-chasing-benchmark-using-64-bit-pointers-in-cuda

ihttps://docs.nvidia.com/launchpad/ai/h100-mig/latest/h100-mig-gpu.html

https://docs.nvidia.com/cuda/cuda-runtime-api/group__CUDART__EXECUTION.html#group__CUDART__EXECUTION_1g6699ca1943ac2655effa0d571b2f4f15

https://developer.download.nvidia.com/compute/DevZone/docs/html/C/doc/html/group__CUDART__HIGHLEVEL_ge0969184de8a5c2d809aa8d7d2425484.html

https://www.stuffedcow.net/research/cudabmk

https://xianweiz.github.io/

https://mediatum.ub.tum.de/doc/1689994/1689994.pdf

HSTU:

https://github.com/facebookresearch/generative-recommenders/pull/72