Three days ago I “upgraded”* to macOS Crapalina 10.15.3.
I would have liked to re-use the figures I obtained in my in-depth review of the 2019 Mac Pro, but when both the OS version changes and Photoshop is updated, the previous data cannot be trusted as valid any longer. I particularly want solid data to comparethe 16-core against the 28-core Mac Pro with Vega II, both of which are now here, though the 16-core must be returned Monday.
I am seeing these losses on the same machine, same memory, same SSD, same script, same everything. Photoshop benchmarks are consistently a little slower to a lot slower, with my most important task of all 15% slower.
Not everything is slower; raw-to-JPEG seems fine and might even be a little faster.
The very long duration update process looks to have involved a firmware update, os presumably it changed stuff yet again in some way that looks to have performance side effects... or it sure seems that way.
I always follow a strict procedure/protocol to get consistent repeatable results, such as waiting for the system to become quiescent before starting a test. So when I test multiple times and keep seeing slower results on more than one test (even if a some tests are only marginally slower), it makes me doubtful that it is a fluke.
Because there are two changes (macOS and Photoshop), I don’t know if macOS is to blame, or Photoshop, or both. But I’d wager that it is macOS Crapalina, which is undoubtedly the most problematic macOS release in history, still in its birthing throes.