ankutalev
ankutalev
@thakkarV thank you, I will check your proposal and response later. But I will be pleased to have answer for 2.x API too, so response from @hwu36 will not be...
@mnicely , yes, sorry for late reply
> Has the hipDNN been discarded? Hello! I'm not a maintainer of this repo and I believe that you already don't need this functionality, but in case if you still...
[BUG][QST] Hopper Grouped GEMM Fails When Workspace not aligned at 64, but MinWorkspaceAlignment =16
cuda-gdb complains for this example built with `-g -G`: ``` Thread 1 "hopper_grouped" received signal CUDA_EXCEPTION_14, Warp Illegal Address. [Switching focus to CUDA kernel 0, grid 4, block (4,0,0), thread...
[BUG][QST] Hopper Grouped GEMM Fails When Workspace not aligned at 64, but MinWorkspaceAlignment =16
@thakkarV can you take a look? Thanks!
[BUG][QST] Hopper Grouped GEMM Fails When Workspace not aligned at 64, but MinWorkspaceAlignment =16
> [@ankutalev](https://github.com/ankutalev), will take a look this week Hi! Any updates?
[BUG][QST] Hopper Grouped GEMM Fails When Workspace not aligned at 64, but MinWorkspaceAlignment =16
@ANIKET-SHIVAM thanks for claryfying! It will be great to have additional assert or changed constant for GroupedGEMM, becasue right now 64bit alignemnt is kind of internal knowledge =) Closing issue,...
> What Haicheng said is far, but please also note that we don't strictly follow semver and have broken compatibility for minor internal methods here and there before. Pretty much...
Hello! I create a [PR ](https://github.com/NVIDIA/cutlass/pull/2457)with this functionality Can you review it, please? cc @mnicely @hwu36
> @ankutalev Thanks for submitting this feature MR. Have you checked the functionality of this feature? Could you post the result of running this feature (example 69) here? Yes, I...