Decision Workspace
atomr-accel-flashattn vs atomr-accel vs atomr-accel-cuda
Side-by-side comparison of Rust crates
FlashAttention v2 + v3 kernel templates for atomr-accel — fp16/bf16/fp8, causal, varlen, ALiBi, sliding window, sink tokens, MQA/GQA, paged KV-cache, and chunked prefill, dispatched through NVRTC + Phase 0.6 cubin cache.
Backend-agnostic compute-acceleration core. Defines the AccelBackend trait, AccelRef<T> typed pointers, AccelError enum, and CompletionStrategy — the abstraction layer that lets atomr-accel-cuda (NVIDIA), and future ROCm / Metal / oneAPI / Vulkan backends plug into the same actor surface.
GPU acceleration via the actor model. Wraps NVIDIA CUDA libraries (cuBLAS, cuDNN, cuFFT, cuRAND, cuSOLVER, cuSPARSE, cuTENSOR, cuBLASLt, NVRTC, NCCL) as supervised atomr actors with generation-validated buffers and a uniform async surface.
Core Metrics
| atomr-accel-flashattn | atomr-accel | atomr-accel-cuda | |
|---|---|---|---|
| Health Score | 52 | 60 | 58 |
| Total Downloads | 44 | 871 | 870 |
| 30d Downloads | 0 | 0 | 0 |
| Dependents | 3 | 29 | 24 |
| Releases | 3 | 4 | 4 |
| Last Updated | 15d ago | 15d ago | 15d ago |
| Age | 19d | 21d | 21d |
Health Breakdown
Technical Details
| atomr-accel-flashattn | atomr-accel | atomr-accel-cuda | |
|---|---|---|---|
| Version | 0.10.0 | 0.10.0 | 0.10.0 |
| Stable (≥1.0) | ✗ No | ✗ No | ✗ No |
| License | Apache-2.0 | Apache-2.0 | Apache-2.0 |
| Dependencies | 11 | 6 | 35 |
| Crate Size | 41KB | 28KB | 319KB |
| Features | 4 | 4 | 54 |
| Yanked % | 0.0% | 0.0% | 0.0% |
| Edition | 2021 | 2021 | 2021 |
| MSRV | 1.78 | 1.78 | 1.78 |
| Owners | 1 | 1 | 1 |
Links
Quick Verdict
- •atomr-accel leads with a health score of 60/100, but none of the options score above 80.