NixOS CUDA | 281 Members | |
| CUDA packages maintenance and support in nixpkgs | https://github.com/orgs/NixOS/projects/27/ | https://nixos.org/manual/nixpkgs/unstable/#cuda | 57 Servers |
| Sender | Message | Time |
|---|---|---|
| 11 Nov 2025 | ||
| here we go again :) | 21:11:13 | |
nix-shell -p --run over nix run -- ? :p | 21:12:32 | |
| Old habits die hard! | 21:13:24 | |
nice | 21:19:28 | |
| lol it's still building jax for whatever reason | 21:27:49 | |
| 🫡 | 21:27:53 | |
master must be ahead of unstable-nixos (the target I think Flox build against) | 21:28:45 | |
| yes. | 21:47:41 | |
We are following nixos-unstable-small on https://cache.nixos.org | 21:48:26 | |
| thanks, is that defined by https://github.com/NixOS/infra/blob/809088c05d04849e3660b22fa9e5bc895570c5fe/channels.nix#L32 ? | 22:16:35 | |
| 12 Nov 2025 | ||
| holy moly. 2tb of ram was not enough for JAX. guess i ran too many jobs at once/ | 03:16:02 | |
seems to have perma-stalled at building python3.12-jax-0.8.0 (pytestCheckPhase): replacing crashed worker gw3 | 03:16:19 | |
| will try again tomorrow :< | 03:29:01 | |
| thanks JAX | 03:29:02 | |
| That entropy ain't gonna' to increase by itself 😜 2048 GiB / 128 (cores) = 16 GiB per core So JAX must need more than that. So even a rig the price of a house can benefit from swap! I | 08:45:31 | |
| * That entropy ain't gonna' to increase by itself 😜 2048 GiB / 128 (cores) = 16 GiB per core So JAX must need more than that. So even a rig the price of a house can benefit from swap! | 08:45:39 | |
| * That entropy ain't gonna' to increase by itself 😜 2048 GiB / 128 cores = 2^11 GiB / 2^7 cores = 2^4 GiB per core = 16 GiB per core So JAX must need more than that. So even a rig the price of a house can benefit from swap! | 08:48:52 | |
| If I remember correctly, I can build jax on 64 cores with 128 GB of RAM | 09:14:48 | |
| * That entropy ain't gonna' increase by itself 😜 2048 GiB / 128 cores = 2^11 GiB / 2^7 cores = 2^4 GiB per core = 16 GiB per core So JAX must need more than that. So even a rig the price of a house can benefit from swap! | 11:33:21 | |
| 128 GiB / 64 cores = 2 GiB per core 🤔 Aha! Must be a quirk with Intel CPUs then! | 12:04:02 | |
| * 128 GiB / 64 cores = 2 GiB per core 🤔 Aha! Must be a quirk with (some, server grade) Intel CPUs then! | 12:05:14 | |
| * 128 GiB / 64 cores = 2 GiB per core 🤔 Aha! Must be a quirk with (some, server grade) Intel CPUs then? | 12:05:26 | |
replacing crashed worker is pytest's message following one of its processes crashing. | 12:12:09 | |
| Yeah, could be for any reason, still seems a bit fishy | 12:16:40 | |
Try with -j 8 for instance. I have never experienced flakiness when building jax, even though I have done it dozens of times. | 12:17:35 | |
| Never tried on Intel though... | 12:17:44 | |
| I also only saw build failures with python3.12 Looks like this particular version is in the team cache, so it built fine
Latest build with the matching hash https://hydra.nixos-cuda.org/build/8123#tabs-buildsteps
🤔 | 12:32:38 | |
| Then maybe some flakiness with your specific CPU... Sometimes things are weird. | 12:37:39 | |
(atlas is configured with cores = 9, maybe this plays a role in building jax successfully) | 12:38:10 | |
| Same exact hashed versions declared from
| 12:38:15 | |