Adobe After Effects Updated With Native Apple Silicon Support, Up to 3x Faster Speeds Than High-End iMac Pro

[MacRumors]

Adobe has updated its professional video editing software After Effects with native M1 support, offering customers up to 3x faster render speeds on Apple's latest Macs compared to high-end Macs with Intel processors.

On M1 computers, Adobe promises up to 2x faster performance in rendering and general app responsiveness. On M1 Ultra, Apple's most high-end chip found in the Mac Studio, Adobe says After Effects will be up to 3x faster for video editors. One specific way Adobe has optimized After Effects is with multi-frame rendering, which utilizes all available cores on Apple silicon to provide a playback experience that's up to 4x faster than a high-end iMac Pro with a 10-core Intel Xeon processor.

The newest version of After Effects will be rolling out to users in the coming days. Adobe also announced several other new features for After Effects and Premiere Pro such as Scene Edit Detection, Auto Color powered by AI, and more.



Article Link: Adobe After Effects Updated With Native Apple Silicon Support, Up to 3x Faster Speeds Than High-End iMac Pro

 

[djcerla]

Some YouTubers returned their Studios a bit too early, it seems.

 

[Notfroggy]

So many benchmarks seem to have had the goal of painting the Apple Silicon Macs in a bad light - usually through disingenuous tactics. I have seen many benchmarks optimised for non-apple API's (openGL anyone?) or comparing non-hardware accelerated against hardware accelerated. for example Handbrake encoding on Intel by default uses hardware acceleration, I have seen benchmarks comparing software rendering on M1 against hardware rendering on Intel/Windows. So much murky stuff designed to confirm a conclusion.

 

[MikeX]

yet the video isn't even 4k. maybe they couldn't export fast enough.

 

[Pressure]

yet the video isn't even 4k. maybe they couldn't export fast enough.

Or YouTube isn't done transcoding the higher resolution video yet ??‍♂️

 

[darthaddie]

So many benchmarks seem to have had the goal of painting the Apple Silicon Macs in a bad light - usually through disingenuous tactics. I have seen many benchmarks optimised for non-apple API's (openGL anyone?) or comparing non-hardware accelerated against hardware accelerated. for example Handbrake encoding on Intel by default uses hardware acceleration, I have seen benchmarks comparing software rendering on M1 against hardware rendering on Intel/Windows. So much murky stuff designed to confirm a conclusion.

Totally agree. it seems like a forever trend to either call new macs mind blowing or “Apple failed again”. Probably that’s what gets them to bait viewers.

Handbrake with software rendering and then Compressor with hardware support are miles apart. Even without yesterdays update compressor beats handbrake on windows (my 12900k and 3090) by a good margin at the same exact settings.

 

[cloudphrenia]

Based on some of the reviews and benchmarks so far, it seems the M1 Max is the best value for performance. The M1 Ultra clearly has more GPU grunt, but not enough to replace workstation GPUs. Either way, these new processors are great additions to the Mac mini lineup.

I would like to see more reviews comparing the M1 Pro to the M1 Ultra. I wonder how far apart the two of them are during heavy workloads.

 

[MauiPa]

I guess when all those “experts” used adobe for benchmarking, and we commented they didn’t know what they were talking about because adobe wasn’t optimized. Well, we were right, those benchmarks were crap

 

[iBluetooth]

There was something odd with all those youtube testers and they getting not the results Apple claimed. But know we get them all redone when the native builds come out.

 

[JippaLippa]

Still not available here in Italy, but I'm beyond excited!
Also most plugins have been made apple silicon native (or are about to), and I believe the After Effects transition will be completed before the end of spring.

 

[aevan]

Based on some of the reviews and benchmarks so far, it seems the M1 Max is the best value for performance. The M1 Ultra clearly has more GPU grunt, but not enough to replace workstation GPUs. Either way, these new processors are great additions to the Mac mini lineup.

They are great replacements for iMac and iMac Pro. They can also compete with Mac Pro for certain workloads. So saying they are great additions to the Mac Mini lineup is a bit downplaying their capabilities.

 

[falkon-engine]

So many benchmarks seem to have had the goal of painting the Apple Silicon Macs in a bad light - usually through disingenuous tactics. I have seen many benchmarks optimised for non-apple API's (openGL anyone?) or comparing non-hardware accelerated against hardware accelerated. for example Handbrake encoding on Intel by default uses hardware acceleration, I have seen benchmarks comparing software rendering on M1 against hardware rendering on Intel/Windows. So much murky stuff designed to confirm a conclusion.

This is disingenuous. The benchmarks you reference were designed at a time when intel processors were in use. Their authors couldn’t have known at development that Apple Silicon was coming. Some just haven’t spent the time necessary to update to their code to natively support Apple Silicon. Could be a cost issue you don’t know the personal details of the developers in question. Perhaps it is non trivial to port their x86 code to ARM. Intel processors can use AVX and other SIMD instructions and even assembly language to access certain functions on the processor that isn’t exposed by the c/c++ language. I had to use assembly language for certain portions of a chess engine I wrote for some speed ups. It is non trivial to port all that code to ARM.

That doesn’t mean that they have malicious intent against Apple and Apple silicon.

 

[Starfyre]

How does it compare to Final Cut Pro X?

 

[adamw]

Wonderful that Adobe updated After Effects. Now if only more software publishers would natively support the full M1 chip family features!

 

[jeffpeng]

Guess a good takeaway from this is to ignore the YouTube cycle of „This Mac beat my 16384835$ 1919 MacPro in this obscure task“ and two weeks later „and here’s why I returned it“.

Basically: nobody want to see a video where the dude says „It’s a good machine, faster doing some, slower doing other, maybe wait for software updates, not a must have but good if you need a new machine.“ Even if that’s what’s actually happening. Typical social media attention w*ing problem.

 

[HiVolt]

Its amazing how long its taking some of the most important software to be Apple Silicon native.

I think I remember the PPC to Intel transition happening much quicker.

 

[killmoms]

Impressive that the move to Apple Silicon unlocks so much performance in After Effects, but boy it's still clearly a single-threaded (or CPU-only) application for a lot of tasks, because the scaling between the M1 Pro and M1 Max is limited. The M1 Ultra is even worse scaling, that's the really surprising one to me. You'd think at this point more plugins/operations would be rearchitected to take advantage of multi-core CPUs and GPUs.

It'd be nice to see more comprehensive AE benchmarks that cover more operations to see which areas are more or less optimized. The M1 Ultra has the potential for amazing performance but it seems like there's still a surprising number of tasks, 15 years into the multi-core–dominant era, that still can't effectively take advantage of the resources on offer.

 

[tubular]

I think I remember the PPC to Intel transition happening much quicker.

I don't think that the PPC to Intel transition involved basically tearing down the graphics capabilities and starting over. The Intel Macs supported OpenGL from the beginning, for example. But the combination of the deprecation of OpenGL, the introduction of Apple Silicon, and the end of the discrete GPU have finally forced major developers to bite the bullet and move to native Metal.

Having said that, it's annoying to see YouTubers saying "oh oh oh Blender on Apple Silicon is so slow now" when it's only officially had Metal support for rendering for about a month.

 

[CoastalMaineBird]

When Adobe puts out videos with words like "s i l i c i o n" on the freaking THUMBNAIL frame, I'm not confident about their attention to detail...

 

[tubular]

When Adobe puts out videos with words like "s i l i c i o n" on the freaking THUMBNAIL frame, I'm not confident about their attention to detail...

Maybe that's why it took so long -- they accidentally ported AE to Apple Silicion first, and then had to go back and port it to Apple Silicon.

 

[Danfango]

So many benchmarks seem to have had the goal of painting the Apple Silicon Macs in a bad light - usually through disingenuous tactics. I have seen many benchmarks optimised for non-apple API's (openGL anyone?) or comparing non-hardware accelerated against hardware accelerated. for example Handbrake encoding on Intel by default uses hardware acceleration, I have seen benchmarks comparing software rendering on M1 against hardware rendering on Intel/Windows. So much murky stuff designed to confirm a conclusion.

If you optimise people's pay checks on social media for attention then you're going to get people saying any old crap for clicks not for facts..

 

[littlebobbytables]

This is disingenuous. The benchmarks you reference were designed at a time when intel processors were in use. Their authors couldn’t have known at development that Apple Silicon was coming. Some just haven’t spent the time necessary to update to their code to natively support Apple Silicon. Could be a cost issue you don’t know the personal details of the developers in question. Perhaps it is non trivial to port their x86 code to ARM. Intel processors can use AVX and other SIMD instructions and even assembly language to access certain functions on the processor that isn’t exposed by the c/c++ language. I had to use assembly language for certain portions of a chess engine I wrote for some speed ups. It is non trivial to port all that code to ARM.

That doesn’t mean that they have malicious intent against Apple and Apple silicon.

You're being disingenuous if you think the original comment was about the developers who made the original benchmarks, not YouTubers who use them maliciously to stir controversy

 

[soyazul]

Based on some of the reviews and benchmarks so far, it seems the M1 Max is the best value for performance. The M1 Ultra clearly has more GPU grunt, but not enough to replace workstation GPUs. Either way, these new processors are great additions to the Mac mini lineup.

I would like to see more reviews comparing the M1 Pro to the M1 Ultra. I wonder how far apart the two of them are during heavy workloads.

those reviews are made over software not yet optimized for m1 Ultra... I dare to say that in less than two years, when every software out there get updated the popular perception will be, the Mac is faster than any desktop windows machine

 

[RedTheReader]

So many benchmarks seem to have had the goal of painting the Apple Silicon Macs in a bad light - usually through disingenuous tactics. I have seen many benchmarks optimised for non-apple API's (openGL anyone?) or comparing non-hardware accelerated against hardware accelerated. for example Handbrake encoding on Intel by default uses hardware acceleration, I have seen benchmarks comparing software rendering on M1 against hardware rendering on Intel/Windows. So much murky stuff designed to confirm a conclusion.

Maybe, just maybe, some of the people comparing Handbrake benchmarks want to use Handbrake on their computers? Maybe some of the people comparing Cinebench scores use Cinema4D for their creative work? Maybe most of the people comparing After Effects render times or Blender render times wanted to run AE or Blender on a daily basis?

Certainly I can’t be the only one here who considered this, right? R… right?

 

[ItsASpider]

"Lol, look now how foolish those reviewers look when they said the Mac wasn't performing well in their reviews."

Do people here know how dumb of an argument this is? Since when does a reviewer review a product as it will be 2 years from that point? That's not what a review is for. M1's performance on many things was, and still is, not that impressive compared to its x86 counterparts. To the end user it doesn't matter if the software got optimized or not or if the poorer performance is due to having to emulate x86, or if it is because it is using a none-Apple API that Apple simply refuses to support for no good reason. If it isn't better, then it just isn't better. That's it. End of story. It doesn't matter if you're given a promis that it will be better 2 years from now. If it isn't good enough for the work someone has to do today, that it just doesn't matter.

When reviewers made these benchmarks in software like After Effects, these were genuin results. They still are.