An Adaptive Pencil Decomposition Library for NVIDIA GPUs
active 2024-07-28 → 2026-07-28 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 731 days from 2024-07-28 to 2026-07-28. Pushes: 142 total, peak 10 in a day. Pull requests: 53 total, peak 5 in a day. Issues: 1 total, peak 1 in a day. Comments: 29 total, peak 4 in a day. Stars: 16 total, peak 2 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| romerojosh | 177 | 121 | 44 | 11 |
| github-actions[bot] | 17 | 0 | 0 | 17 |
| p-costa | 1 | 0 | 0 | 1 |
| fallintoplace | 1 | 0 | 1 | 0 |
Recent activity
Latest issues, pull requests and releases
- Pull request#146fallintoplace2026-06-24 17:29
- Issue comment#146github-actions[bot]2026-06-24 17:25Fix direct transpose output offset
- Pull request#134romerojosh2026-06-04 20:47
- Pull request#131romerojosh2026-05-06 23:11
- Pull request#129romerojosh2026-05-05 21:17
- Pull request#127romerojosh2026-05-04 21:42
- Pull request#126romerojosh2026-05-04 18:44
- Issue comment#124github-actions[bot]2026-04-30 23:16Use ref-count based handling of NVSHMEM initialization state.
- Pull request#122romerojosh2026-04-30 17:41
- Issue comment#121github-actions[bot]2026-04-29 22:09Make header dependencies explicit.
- Issue comment#121github-actions[bot]2026-04-29 22:03Make header dependencies explicit.
- Issue comment#114github-actions[bot]2026-03-13 20:48Add new NVSHMEM transpose communication backend with SM-based P2P copies.
- Issue comment#109github-actions[bot]2026-03-12 23:08Adjust workspace sizing to allow 256-byte alignment of workspace offset pointers.
- Issue#112romerojosh2026-03-11 22:21cuTENSOR large-tensor permutation bug causes heap corruption in cudecompTranspose for tensors with >~1B elements
- Issue comment#108github-actions[bot]2026-03-04 21:07Force workspace usage with MPI backends on MNNVL communicators when fabric allocated workspace is available.
- Issue comment#108romerojosh2026-03-04 20:27Disable transpose shortcut paths for MPI backends on MNNVL communicators when fabric allocated workspace is available.
- Issue comment#107github-actions[bot]2026-03-04 18:49Use native NVSHMEM synchronization APIs in NVSHMEM backends
- Pull request#106romerojosh2026-02-23 21:57
- Issue comment#106github-actions[bot]2026-02-21 01:03Enforce NVSHMEM minimum version of 2.6.0 to remove old workarounds. Update NVSHMEM usage guidance.
- Issue comment#106romerojosh2026-02-21 01:03Enforce NVSHMEM minimum version of 2.6.0 to remove old workarounds. Update NVSHMEM usage guidance.
- Pull request#105romerojosh2026-02-17 23:10
- Issue comment#102romerojosh2026-02-11 00:35Enabling support for process decompositions with empty pencils.
- Pull request#102romerojosh2026-02-10 20:54
- Issue comment#100github-actions[bot]2026-02-04 17:160.6.1 release
- Issue comment#99github-actions[bot]2026-01-28 21:05Add CUDECOMP_DISABLE_MNNVL debug option.
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 16 stars here means stars gained during the window, not the repo's star count.