Skip to content

{2026.06} CUDA 13.3.0 and cuDNN 9.23.0.39 - #1567

Draft
bedroge wants to merge 2 commits into
EESSI:mainfrom
bedroge:2026.06_cuda_cudnn
Draft

{2026.06} CUDA 13.3.0 and cuDNN 9.23.0.39#1567
bedroge wants to merge 2 commits into
EESSI:mainfrom
bedroge:2026.06_cuda_cudnn

Conversation

@bedroge

@bedroge bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator

Draft, as we need to agree on the supported GPU targets for 2026.06 first.

@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

bot: build repo:eessi.io-2026.06-software instance:eessi-bot-rug for:arch=x86_64/amd/zen5,accel=nvidia/cc120

@eessi-bot-rug

eessi-bot-rug Bot commented Aug 11, 2026

Copy link
Copy Markdown

New job on instance eessi-bot-rug for repository eessi.io-2026.06-software
Building on: amd-zen5 and accelerator nvidia/cc120
Building for: x86_64/amd/zen5 and accelerator nvidia/cc120
Job dir: /scratch/hb-eessibot/SHARED/jobs/2026.08/pr_1567/30567433

date job status comment
Aug 11 09:47:42 UTC 2026 submitted job id 30567433 awaits release by job manager
Aug 11 09:49:35 UTC 2026 released job awaits launch by Slurm scheduler
Aug 11 09:51:39 UTC 2026 running job 30567433 is running
Aug 11 10:05:54 UTC 2026 finished
😢 FAILURE (click triangle for details)
Details
✅ job output file slurm-30567433.out
✅ no message matching FATAL:
❌ found message matching ERROR:
❌ found message matching FAILED:
❌ found message matching required modules missing:
❌ no message matching No missing installations
✅ found message matching .tar.* created!
Artefacts
eessi-2026.06-software-linux-x86_64-amd-zen5-accel-nvidia-cc120-17864426840.tar.zstsize: 2246 MiB (2355473685 bytes)
entries: 10063
modules under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/reprod
CUDA/13.3.0/20260811_100423UTC
other under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120
no other files in tarball
Aug 11 10:05:54 UTC 2026 test result
😢 FAILURE (click triangle for details)
Reason
EESSI test suite was not run, test step itself failed to execute.
Details
✅ job output file slurm-30567433.out
❌ found message matching ERROR:
✅ no message matching [\s*FAILED\s*].*Ran .* test case

@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

bot: build repo:eessi.io-2026.06-software instance:eessi-bot-rug for:arch=x86_64/amd/zen5,accel=nvidia/cc120

@eessi-bot-rug

eessi-bot-rug Bot commented Aug 11, 2026

Copy link
Copy Markdown

New job on instance eessi-bot-rug for repository eessi.io-2026.06-software
Building on: amd-zen5 and accelerator nvidia/cc120
Building for: x86_64/amd/zen5 and accelerator nvidia/cc120
Job dir: /scratch/hb-eessibot/SHARED/jobs/2026.08/pr_1567/30570957

date job status comment
Aug 11 10:13:26 UTC 2026 submitted job id 30570957 awaits release by job manager
Aug 11 10:13:59 UTC 2026 released job awaits launch by Slurm scheduler
Aug 11 10:16:03 UTC 2026 running job 30570957 is running
Aug 11 10:18:05 UTC 2026 finished
😢 FAILURE (click triangle for details)
Details
✅ job output file slurm-30570957.out
✅ no message matching FATAL:
❌ found message matching ERROR:
❌ found message matching FAILED:
❌ found message matching required modules missing:
❌ no message matching No missing installations
✅ found message matching .tar.* created!
Artefacts
eessi-2026.06-software-linux-x86_64-amd-zen5-accel-nvidia-cc120-17864433290.tar.zstsize: 2246 MiB (2355473610 bytes)
entries: 10063
modules under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120/reprod
CUDA/13.3.0/20260811_101515UTC
other under 2026.06/software/linux/x86_64/amd/zen5/accel/nvidia/cc120
no other files in tarball
Aug 11 10:18:05 UTC 2026 test result
😢 FAILURE (click triangle for details)
Reason
EESSI test suite was not run, test step itself failed to execute.
Details
✅ job output file slurm-30570957.out
❌ found message matching ERROR:
✅ no message matching [\s*FAILED\s*].*Ran .* test case

@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

bot: build repo:eessi.io-2026.06-software instance:eessi-bot-surf for:arch=x86_64/intel/icelake,accel=nvidia/cc80

edit: surf bot is offline

@EESSI EESSI deleted a comment from eessi-bot-aws Bot Aug 11, 2026
@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

bot: build repo:eessi.io-2026.06-software instance:eessi-bot-mc-aws on:arch=aarch64/generic for:arch=aarch64/generic,accel=nvidia/cc75+nvidia/cc80+nvidia/cc90+nvidia/cc100+nvidia/cc120

@eessi-bot-aws

eessi-bot-aws Bot commented Aug 11, 2026

Copy link
Copy Markdown

New job on instance eessi-bot-mc-aws for repository eessi.io-2026.06-software
Building on: generic
Building for: aarch64/generic and accelerator nvidia/cc75+nvidia/cc80+nvidia/cc90+nvidia/cc100+nvidia/cc120
Job dir: /project/def-users/SHARED/jobs/2026.08/pr_1567/185690

date job status comment
Aug 11 10:22:09 UTC 2026 submitted job id 185690 awaits release by job manager
Aug 11 10:22:39 UTC 2026 released job awaits launch by Slurm scheduler
Aug 11 10:23:43 UTC 2026 running job 185690 is running
Aug 11 11:11:00 UTC 2026 finished
🤷 UNKNOWN (click triangle for detailed information)
  • Job results file _bot_job185690.result does not exist in job directory or reading it failed.
  • No artefacts were found/reported.
Aug 11 11:11:00 UTC 2026 test result
🤷 UNKNOWN (click triangle for detailed information)
  • Job test file _bot_job185690.test does not exist in job directory or reading it failed.

@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

bot: build repo:eessi.io-2026.06-software instance:eessi-bot-mc-aws on:arch=aarch64/generic for:arch=aarch64/generic,accel=nvidia/cc75+nvidia/cc80+nvidia/cc90+nvidia/cc100+nvidia/cc120

@eessi-bot-aws

eessi-bot-aws Bot commented Aug 11, 2026

Copy link
Copy Markdown

New job on instance eessi-bot-mc-aws for repository eessi.io-2026.06-software
Building on: generic
Building for: aarch64/generic and accelerator nvidia/cc75+nvidia/cc80+nvidia/cc90+nvidia/cc100+nvidia/cc120
Job dir: /project/def-users/SHARED/jobs/2026.08/pr_1567/185732

date job status comment
Aug 11 14:04:49 UTC 2026 submitted job id 185732 awaits release by job manager
Aug 11 14:05:13 UTC 2026 released job awaits launch by Slurm scheduler
Aug 11 14:06:20 UTC 2026 running job 185732 is running
Aug 11 14:44:40 UTC 2026 finished
😢 FAILURE (click triangle for details)
Details
✅ job output file slurm-185732.out
✅ no message matching FATAL:
❌ found message matching ERROR:
❌ found message matching FAILED:
❌ found message matching required modules missing:
❌ no message matching No missing installations
✅ found message matching .tar.* created!
Artefacts
eessi-2026.06-software-linux-aarch64-generic-accel-nvidia-cc75-accel-nvidia-cc80-accel-nvidia-cc90-accel-nvidia-cc100-accel-nvidia-cc120-17864588030.tar.zstsize: 12471 MiB (13077209597 bytes)
entries: 89665
modules under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc75/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc75/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc75/reprod
CUDA/13.3.0/20260811_141004UTC
other under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc75
no other files in tarball
modules under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc80/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc80/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc80/reprod
CUDA/13.3.0/20260811_141342UTC
other under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc80
no other files in tarball
modules under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc90/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc90/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc90/reprod
CUDA/13.3.0/20260811_141741UTC
other under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc90
no other files in tarball
modules under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc100/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc100/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc100/reprod
CUDA/13.3.0/20260811_142324UTC
other under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc100
no other files in tarball
modules under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc120/modules/all
CUDA/13.3.0.lua
software under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc120/software
CUDA/13.3.0
reprod directories under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc120/reprod
CUDA/13.3.0/20260811_142926UTC
other under 2026.06/software/linux/aarch64/generic/accel/nvidia/cc120
no other files in tarball
Aug 11 14:44:40 UTC 2026 test result
😢 FAILURE (click triangle for details)
Reason
EESSI test suite was not run, test step itself failed to execute.
Details
✅ job output file slurm-185732.out
❌ found message matching ERROR:
✅ no message matching [\s*FAILED\s*].*Ran .* test case

@bedroge

bedroge commented Aug 11, 2026

Copy link
Copy Markdown
Collaborator Author

This is weird, cuDNN is complaining that cuobjdump cannot be found:

ERROR: Installation of cuDNN-9.23.0.39-CUDA-13.3.0.eb failed: Failed to execute CUDA sanity check: cuobjdump not found

It's part of the CUDA 13.3.0 installation though, and that succeeded. Going to try this interactively to figure out what's wrong here.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant