Is this a duplicate?
Area
cuda.core
Is your feature request related to a problem? Please describe.
Thus far, the approach with skipping tests on unsupported hardware on cuda.core.system has been to just catch a NotSupportedError and then skip the test. For APIs where we know the minimum supported hardware rev, we should instead test on that. This will be more robust to tests failing where we don't expect them to.
Describe the solution you'd like
For each test that uses non-universally supported APIs, check the value of device_architecture.id, and skip the test based on that.
Describe alternatives you've considered
No response
Additional context
No response