Skip to content

Describe a backend by what it can do, not by its name (#151) #151

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
joy-xiaojizhang wants to merge 1 commit into from
Closed

Describe a backend by what it can do, not by its name (#151) #151

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
13 changes: 13 additions & 0 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -226,6 +226,19 @@ KernelAgent supports multiple GPU platforms for Triton kernel execution:
|----------|---------------|------|--------|
| NVIDIA CUDA | `cuda` | `--target-platform cuda` (default) | Fully supported |
| Intel XPU | `xpu` | `--target-platform xpu` | Supported |
| Fake (no accelerator) | `cpu` | `--target-platform fake` | Dry runs and CI without hardware |

A backend is described by the capabilities it declares — availability check,
device setup, synchronization, test prelude and default worker count — rather
than by its name, so nothing in the generic templates branches on a device
string. Register one with
`triton_kernel_agent.platform_config.register_platform`, and register its
optimization components with
`triton_kernel_agent.platform.registry.registry.register`.

The `fake` backend exists to exercise the pipeline where there is no
accelerator. It reports a placeholder time so the pipeline can complete, and
nothing it produces is a performance claim.

### Intel XPU Notes

Expand Down
Loading