Decision Workspace
fandhe-ai-backend-cuda vs burn-cuda vs baracuda-cutlass
Side-by-side comparison of Rust crates
49
fandhe-ai-backend-cuda
experimentalv0.11.0
CUDA backend (dynamic-loading via cudarc) for a from-scratch Rust AI/ML library
74
burn-cuda
growingv0.22.0
CUDA backend for the Burn framework
59
baracuda-cutlass
experimentalv0.0.1-alpha.86
Safe Rust wrapper for compiled CUTLASS kernels: plan-based GEMM and grouped GEMM with caller-supplied workspace, typed device-buffer arguments, and capture-safe launch.
Core Metrics
| fandhe-ai-backend-cuda | burn-cuda | baracuda-cutlass | |
|---|---|---|---|
| Health Score | 49 | 74 | 59 |
| Total Downloads | 378 | 1.3M | 3.1K |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 9 | 368 | 55 |
| Releases | 9 | 28 | 58 |
| Last Updated | today | 3d ago | 3d ago |
| Age | 1m | 2y 1m | 5m |
Health Breakdown
fandhe-ai-backend-cuda
Maintenance
14
Quality
13
Community
9
Popularity
3
Documentation
10
burn-cuda
Maintenance
20
Quality
17
Community
17
Popularity
7
Documentation
13
baracuda-cutlass
Maintenance
19
Quality
15
Community
11
Popularity
4
Documentation
10
Technical Details
| fandhe-ai-backend-cuda | burn-cuda | baracuda-cutlass | |
|---|---|---|---|
| Version | 0.11.0 | 0.22.0 | 0.0.1-alpha.86 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | MIT OR Apache-2.0 | MIT OR Apache-2.0 | MIT OR Apache-2.0 |
| Dependencies | 3 | 4 | 10 |
| Crate Size | 1.8MB | 37KB | 78KB |
| Features | 1 | 7 | 5 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2024 | 2024 | 2024 |
| MSRV | — | — | — |
| Owners | 1 | 3 (team) | 1 |
Links
Quick Verdict
- •burn-cuda leads with a health score of 74/100, but none of the options score above 80.
- •burn-cuda is depended on by 368 crates — strongest ecosystem trust.