Tweet from OTOY - “Octane X on this #MacPro - vs. the BOXX PC tested by
@Verge below - could end up being faster on the MacPro (as measured in standard OctaneBench 4 test). Subject to change & based on some VERY early testing we have done internally. That’s with RTX off of course!“
I wonder if this is Metal vs CUDA optimization or just the Vega II’s?
Man, Twitter is just soooo annoying. Why in 2020 do people think the character limit is cool? Because I felt like I was trying to interpret hieroglyphics just to find a link to a review. But I digress.
The crux of the review to me was two-fold: First, and foremost, Adobe's apps still just suck. They try to put that on Adobe needing to optimize for the Mac Pro, and excuse Adobe because it's so new but that's giving Adobe a huge pass. High-end hardware has been around for years, and Adobe has done little to *nothing* to take advantage of it. IMO, the subscription model promised more continuous investment by Adobe because of continuous revenue, but instead led to them just sitting back and taking payments without any need to invest in anything, because as a consumer you're trapped. Apple has no responsibility for how absolutely terribly their apps run, because that's not just a Mac/Mac Pro issue. It's an Adobe issue.
Second, it's a pro machine, and designed for heavy workloads. A bunch of the folks they gave it to do pretty light levels of work, and surprise, surprise they didn't see much benefit. The one quote that made me literally laugh out loud? "I was working off a server via VPN, so it’s possible that was slowing things down." Ummm, yes!
On their benchmark side...hard to really assess what was going on there. The Xeon is not an ideal processor for some of those types of test/activities. The bulk of the tests are Adobe Premiere related, which we know is an issue--and interestingly enough there's not a compelling advantage to either machine in those results--some swing one way, some the other. More an Adobe issue again, no matter what hardware you buy. The Cinebench score is not surprising, that's an area that the Ryzen is known to do well and there's no sense rehashing the limitations of that as a useful benchmark for anything besides Cinebench. It does seem to address your question that there is a question on whether Apple/AMD will be able to get better performance with the Vega II's with better drivers, or whether they just don't have the power to match the Nvidia cards; the Vega II is based on an older architecture.