Decision Workspace
oxicuda-train vs oxicuda-infer vs burn-cuda
Side-by-side comparison of Rust crates
43
oxicuda-train
experimentalv0.1.8
GPU-accelerated training engine: fused optimizer update kernels, gradient checkpointing, LR schedulers, and ZeRO optimizer state sharding
44
oxicuda-infer
experimentalv0.1.8
High-performance GPU inference engine for OxiCUDA: PagedAttention KV cache, continuous batching, speculative decoding, structured sampling, and quantized inference.
68
burn-cuda
growingv0.21.0
CUDA backend for the Burn framework
Core Metrics
| oxicuda-train | oxicuda-infer | burn-cuda | |
|---|---|---|---|
| Health Score | 43 | 44 | 68 |
| Total Downloads | 116 | 126 | 584.5K |
| 30d Downloads | 0 | 0 | 90.9K |
| Dependents | 0 | 0 | 263 |
| Releases | 9 | 9 | 23 |
| Last Updated | 4d ago | 4d ago | 18d ago |
| Age | 1m | 1m | 1y 9m |
Health Breakdown
oxicuda-train
Maintenance
14
Quality
9
Community
6
Popularity
2
Documentation
12
oxicuda-infer
Maintenance
14
Quality
9
Community
6
Popularity
3
Documentation
12
burn-cuda
Maintenance
19
Quality
17
Community
12
Popularity
7
Documentation
13
Technical Details
| oxicuda-train | oxicuda-infer | burn-cuda | |
|---|---|---|---|
| Version | 0.1.8 | 0.1.8 | 0.21.0 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | Apache-2.0 | Apache-2.0 | MIT OR Apache-2.0 |
| Dependencies | 9 | 3 | 4 |
| Crate Size | 68KB | 72KB | 30KB |
| Features | 2 | 0 | 8 |
| Yanked % | 22.2% | 22.2% | 0.0% |
| Edition | 2024 | 2024 | 2024 |
| MSRV | 1.85 | 1.85 | — |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •burn-cuda leads with a health score of 68/100, but none of the options score above 80.
- •burn-cuda is depended on by 263 crates — strongest ecosystem trust.