Open
Description
We calibrated the render pass counting performance tests that run on CI to the speed of the emulators on CI that we use. However, if we re-run locally the outcomes change and its confusing as to why. We need to make it very clear in those test messages the scenarios that are compute dependent. We should also consider just not including them in CI testing.