chrissantoss/Optimized-Sparse-Attention-Kernel-for-Simulating-Cosmic-Structures
View on GitHub ↗Related repositories →CUDA kernel for sparse attention, optimized for cosmic simulations (e.g., galaxy formation) using CUTLASS, Tensor Cores, and Nsight. Integrated into JAX via pybind11. Reduces memory 4x, speeds up 2-3x vs. dense attention. Demo simulates N=4096 particles, aligning with xAI’s mission to accelerate scientific discovery.
active 2025-03-09 → 2025-03-10 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 2 days from 2025-03-09 to 2025-03-10. Pushes: 3 total, peak 3 in a day. Pull requests: 0 total, peak 0 in a day. Issues: 0 total, peak 0 in a day. Comments: 0 total, peak 0 in a day. Stars: 0 total, peak 0 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| chrissantoss | 3 | 3 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
No issue or PR events — this repo's activity is pushes only.
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 0 stars here means stars gained during the window, not the repo's star count.