Bought Atmos only Mac last December, now have to pay for it!?

[Darryl Ramm]

Originally Posted by atticmike

Since this test was done prior to native compatibility, just add another 10% to the Mac Studio and you know where it is performance wise.

Some good points, and that's an interesting test shown, but you can't just add 10% and assume that is what you get natively. For high-end system running under real load you need to repeat the benchmark natively. It might be a lot better, it might not, it might even in pathological cases be worse (but very unlikely not). When pushing these limits tuning, optimization, plugins etc. can just be critical. And binary translation software can have some pathological corner cases so maybe dangerous to assume anything when pushing limits. These guys are good, much better than some other folks who talk about performance. Would be interesting to know all the other optimizations/config variations they did, how things work at 1k buffers etc.

My broad argument would be for many things x86-64 processors that most folks are likely to buy I'd expect a closer to parity performance. I reacted to "Heavily inferior" at least to me sounds like a very large difference in performance, maybe the difference between usable and not or impressive and underwhelming.

--

For those other multi-core i9 comparisons, I'd forgot how well the 13th gen did, especially with CineBench. And we are partially here with Cinebench being used so much because AMD helped pushed Cinebench scoring with their superior processors for a while, and now Intel is competitive Cinebench is a natural comparison to want to make, and I bet some AMD folks are smarting about being out Cinebenched. While all these benchmarks don't make up for real world testing with your own workload they do show that overall we live in pretty good times with modern processing horsepower.

Now not wanting to be mistaken for an Apple fan boi, Apple does need to deliver M3 sooner than later, they had a great story esp. with performance/power but their supply chain problems and miserly M2 update are not great. And along comes an new iPhone gen that will skew priorities. And if M3 ends up all next year, that's not good.

[atticmike]

Originally Posted by Darryl Ramm

Some good points, and that's an interesting test shown, but you can't just add 10% and assume that is what you get natively. For high-end system running under real load you need to repeat the benchmark natively. It might be a lot better, it might not, it might even in pathological cases be worse (but very unlikely not). When pushing these limits tuning, optimization, plugins etc. can just be critical. And binary translation software can have some pathological corner cases so maybe dangerous to assume anything when pushing limits. These guys are good, much better than some other folks who talk about performance. Would be interesting to know all the other optimizations/config variations they did, how things work at 1k buffers etc.

My broad argument would be for many things x86-64 processors that most folks are likely to buy I'd expect a closer to parity performance. I reacted to "Heavily inferior" at least to me sounds like a very large difference in performance, maybe the difference between usable and not or impressive and underwhelming.

--

For those other multi-core i9 comparisons, I'd forgot how well the 13th gen did, especially with CineBench. And we are partially here with Cinebench being used so much because AMD helped pushed Cinebench scoring with their superior processors for a while, and now Intel is competitive Cinebench is a natural comparison to want to make, and I bet some AMD folks are smarting about being out Cinebenched. While all these benchmarks don't make up for real world testing with your own workload they do show that overall we live in pretty good times with modern processing horsepower.

Now not wanting to be mistaken for an Apple fan boi, Apple does need to deliver M3 sooner than later, they had a great story esp. with performance/power but their supply chain problems and miserly M2 update are not great. And along comes an new iPhone gen that will skew priorities. And if M3 ends up all next year, that's not good.

I have an M1 MacBook Air for on the road and I couldn't be happier, there is where the apple silicone for me as an audio engineer shines but as a main mix / production Workstation really not. My main rig is a 13600k with 128 gb of ram and I don't think that will change back to Apple since 2018 any time soon and I'm happy that Pro Tools runs now flawlessly on Windows without freezes, errors or playback interruptions. I recently upgraded from a 9900k and it was actually not necessary given that 250 track sessions with plenty of plugins rarely exceeded the 50% mark, now I'm at 20%

I mean generally, ARM could probably benefit from plugin manufacturers' optimizations which might hold a bunch of them back even natively in performance but that will take time but ARM will never take off in audio heavy multi-threading because x86 in terms of architecture is superior in parallel processing unless Apple utilizes GPU processing in combination of the CPU.