Update Buildkite pipeline for the new JuliaGPU cluster

[maleadt]

The JuliaGPU Buildkite agents have moved to a dedicated cluster, where the single juliagpu queue has been split into per-backend queues. Steps now select agents using queue: "cuda", queue: "rocm" or queue: "oneapi" instead of queue: "juliagpu" combined with a cuda/rocm/intel tag. The pipeline-upload trigger steps in pipeline.yml, which previously targeted the generic juliagpu queue without a backend tag, now run on the cuda queue. The commented-out AMDGPU group was updated as well.

This change only affects agent selection; the steps themselves are unchanged.

🤖 Generated with Claude Code

[gemini-code-assist]

Code Review

This pull request updates the Buildkite pipeline configurations to use the new "cuda" and "rocm" queues instead of "juliagpu". The reviewer recommends using a lightweight CPU queue instead of the "cuda" queue for pipeline upload steps to avoid occupying valuable GPU resources for non-GPU tasks.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.

[gemini-code-assist]

Running the pipeline upload step on the "cuda" queue occupies a GPU agent for a non-GPU task, which can lead to inefficient resource utilization and longer wait times for actual GPU tests. If a lightweight or default CPU queue is available in the new cluster, using it for pipeline uploads is highly recommended.