Mmm. My concern with the Iray renderer is that the CPU rendering code gets generated ‘on-the-fly’ when the render starts, so it wouldn’t benefit from the same install-time transpilation that you’re seeing; it’s more likely to use the just-in-time features of Rosetta2, so there’ll be some kind of a performance hit when the render kicks off. But so long as it can do some optimization, it might not be too bad. I certainly hear that performance is respectable, even if it can’t use the full power of the silicon in the most efficient way.
Apple’s ability to support old code on new CPUs is impressive: they’re now on their third architecture transition (68K to PowerPC, PowerPC to Intel, Intel to Apple Silicon) and each time they’ve managed to keep older software going on the new architecture, at least for a little while. Ironically, where they fall down is on software transitions: I can run software written for Windows 7 flawlessly on Windows 11, but every time Apple ships a new OS, a few more older apps stop working.
