Heterogeneous Devices
GPUStack makes GPUs, NPUs, MLUs, DCUs and PPUs available to workloads. The manufacturer controls the resource name and runtime isolation method.
Contents
Overview
Device Manager detects hardware and publishes capacity. AcceleratorType selects accelerated pools
for serving and prefetch; InstanceType sizes interactive workspaces. Both share the canonical pool
ClusterQueue. Kueue admits queued requests, and the device plugin allocates devices to containers.
Configuration
Accelerator requests
Choose a resource family in Accelerator Requests . The guide includes request examples and validation rules. Device Discovery explains the inventory and allocation boundary.
Operations
Device operations
- Step 1: Check the node. Run Preflight Checks before installation.
- Step 2: Request devices. Follow the device walkthrough .
- Step 3: Verify allocation. Check the
Workload,Podand device inventory in the walkthrough.
For physical partitions, use the NVIDIA , T-Head , or Hygon runbook. The scheduling chain describes queue status.
Troubleshooting
For a rejected request, check the admission gates and request rules. For node checks and their results, see Preflight Checks . For partition recovery, use the runbook for the manufacturer.