Open code, practice, and notes →
Download source with English comments · for reading
WebGPU · 22 Device Loss and Reinitialization
Series contents · Technology map
Stage: Recovery, Measurement, and an Integrated Project · Editable example · Storage ID: webgpu-step-22
Goals and learning outcomes
- Handle device loss.
- Validate the resource recreation path.
Prerequisites
Continue using the output and notes from unit 21. Preserve the previous inputs and reference outputs before the Device Loss and Reinitialization experiment.
Key concepts and explanations
Lost State
device.lost signals that the device is no longer valid. Resources from the previous device cannot be reused unchanged on a new device.
Recreation
Recovery needs a path that recreates shaders, buffers, and pipelines. Retain CPU settings and initial data in a reproducible form.
Practice setup
First read the execution-scope guidance on the example screen. The following preparation is for extending the assignment in a separate project. This is a document-guided exercise to implement yourself in a separate localhost or HTTPS project. Check the secure context, navigator.gpu, and adapter request result, and provide an unsupported-environment message. Preserve device.lost handling and resource release.
The linked example is a starting point. It does not implement every task below. Complete work beyond its supported scope in a separate project.
Practice tasks
- In device.js, connect device.lost handling to stopping the render loop.
- Call device.destroy from a development button and display the loss reason.
- On explicit user retry, create a new device and resources and restore saved settings.
Variables to observe
| Variable | Baseline | Change | Expected change / hypothesis |
|---|---|---|---|
| Device state | Normal | Intentional loss through destroy | The loop is expected to stop and recover with new resources after an explicit retry. |
Predictions in this table are not measured results. Change one condition at a time and record why the actual result differs.
Completion criteria
- Do not submit to the previous device after loss.
- After retry, create the same initial scene with a new device.
Self-check
- Should retry policies for a destroy reason and unexpected loss be identical?
- What information must remain on the CPU for recovery?
Official references
Linked example and execution scope
Example source · Runtime environments and limits
Runs real WebGPU/WGSL. Requires HTTPS/localhost and a supported adapter; unsupported environments display a reason, and CPU results are not substituted for GPU results.
Source code and records are not overwritten automatically. Resetting an existing example also preserves your notes.
Next learning
Previous: Instancing and Data Structures · Next: Asynchronous Errors and Task Identity