> **tensor4all temporary fork.** This `t4a-cubecl*` package is a temporary tensor4all fork used by `tenferro-rs` while required patches are being upstreamed to the official CubeCL project. It is not a replacement for upstream CubeCL.
# CubeCL Runtime
This crate helps creating high performance async runtimes.
- [x] Asynchronous kernel executions
- [x] Memory allocation management
- [x] Autotuning