Translating an x86 application using Rosetta 2 to run on Apple Silicon chips will reduce performance. However, as Apple showed at WWDC when demonstrating Maya and Shadow of the Tomb Raider on an A12Z Mac mini developer kit, switching to Apple Silicon may be a net performance increase on some models.
We do not know what kind of performance the first Apple Silicon chips will have, but at the beginning of the PowerPC to Intel transition, the developer kits had Intel chips much slower than the ones in the first Intel macs. Also, Apple wants developers to test their applications to run well even on systems with less performance. So, I anticipate the first retail chips will be faster than the A12Z chip in the developer kit.
Setting aside bugs, if the first retail Apple Silicon chips are as fast as the A12Z, which current Intel models may run native x86 applications slower than the first Apple Silicon models can run them as translated applications?
We do not know what kind of performance the first Apple Silicon chips will have, but at the beginning of the PowerPC to Intel transition, the developer kits had Intel chips much slower than the ones in the first Intel macs. Also, Apple wants developers to test their applications to run well even on systems with less performance. So, I anticipate the first retail chips will be faster than the A12Z chip in the developer kit.
Setting aside bugs, if the first retail Apple Silicon chips are as fast as the A12Z, which current Intel models may run native x86 applications slower than the first Apple Silicon models can run them as translated applications?