Skip to content
This repository has been archived by the owner on Jun 27, 2024. It is now read-only.

Issues: iree-org/iree-nvgpu

13 Open 3 Closed
13 Open 3 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Issues list

[Triton] Compute capability for hal.executable.source must match IREE cuda SM version bug Something isn't working enhancement New feature or request
#144 opened Jun 15, 2023 by ezhulenev
[Triton] Construct stream.executable from Triton functions enhancement New feature or request
#143 opened Jun 15, 2023 by ezhulenev
[Triton] Use nested passes to lower Triton to HAL executable enhancement New feature or request
#142 opened Jun 15, 2023 by ezhulenev
[cuDNN] Use semaphores to synchronize before/after cuDNN operations enhancement New feature or request
#139 opened Jun 14, 2023 by ezhulenev
[cuDNN] cuDNN module should create a dedicated CUDA stream enhancement New feature or request
#138 opened Jun 14, 2023 by ezhulenev
[Triton] Add minimal matmul example (no auto tuning) enhancement New feature or request
#129 opened Jun 9, 2023 by ezhulenev
[Triton] Triton executables should support shared memory enhancement New feature or request
#128 opened Jun 9, 2023 by ezhulenev
[Triton] Do not use temp files for passing PTX to HAL executable enhancement New feature or request
#126 opened Jun 9, 2023 by ezhulenev
[Triton] Block dimension should be inferred from the Triton function enhancement New feature or request
#125 opened Jun 9, 2023 by ezhulenev
[Triton] num-warps and num-stages should be a property of triton.executable.export enhancement New feature or request
#124 opened Jun 9, 2023 by ezhulenev
Set up continuous integration for OpenXLA Nvgpu project enhancement New feature or request
#58 opened May 4, 2023 by ezhulenev
@frgossen
3
[RFC] First class Triton support in OpenXLA Nvgpu
#54 opened May 4, 2023 by ezhulenev
6
[RFC] Integration with cuDNN via IREE compiler/runtime plugins enhancement New feature or request
#12 opened Apr 18, 2023 by ezhulenev
@ezhulenev @frgossen @chsigg
1
ProTip! Type g i on any issue or pull request to go back to the issue listing page.