Compute-GPU-Profile: Unified performance analysis CLI for scalar, SIMD, wgpu, and CUDA workloads
Part of the [Aprender](https://github.com/paiml/aprender) monorepo — 70 workspace crates.
```bash
cargo install aprender # CLI binary
```
```toml
[dependencies]
aprender-cgp = "0.29"
```
- -