OF 3D Frame rate in m1 Mac, It's Awesome ~!

It’s a 3D graphics test (render, grid, texture, rotation…)

The screen frame rate is 1190 frames/sec! … that’s shocking.
It comes out about ~90 frame/sec on My Mac.

It would be great if it was run with Rosetta 2 (a feature that emulates an Intel Mac app running on arm core) that is not native.

How fast will it be if you run it native?

Probably like the Envidia Jetson series
It seems to be because of the UMA memory that integrates system memory and graphic memory…

Of course, it is said that CPU/GPU is also fast, but I didn’t know that it was that much.



1 Like


Thanks for the post.

I should say that the example was running with vertical sync off. So it might not be as massive a difference.

Try on your Intel setup with:

in the ofApp::setup() function.

Instead of ofSetVerticalSync(true);

I recently got the same example building natively on the M1 Mini and it was a similar fps.
Which just might mean we aren’t taxing the CPU/GPU very much with this example.

I’ll try and find a way to highlight a more meaningful difference between Intel, M1 Rosetta and M1 Native setups.


I tried the same 3D primitive example with the resolution on the highest ( press key 4 ):

  • M1 Mac Mini ( Running OF Native ): 1244 fps
  • Intel Macbook Pro 2.9 GHz w/Radeon Pro 560X : 870 fps

So that’s pretty impressive!



Ho, @theo

That’s Great ~!

Is Native OF already compiled ?



@bemoregt still working on getting all the libs building nicely and need to update the Xcode project template but it is close to be being fully compatible.

With the core libs compiled so far I was able to get the 3DPrimitivesExample working but it will be a bit more work to get an official release out with the changes.

I’ll keep this thread up to date as we get closer.



  • verticalsync= true

  • verticalsync= false

1050 frame/sec on my MacPro2013. )-:

  • Intel Xeon 3.3GHz 6-cores
  • 32GB RAM ECC
  • Radeon X580 8GHz
  • NVMe 250GB
  • OF v0.11.0_OSX
  • OSX Mojave


Try pressing 4 on your keyboard with verticalsync off I am guessing it will drop a fair amount below 1050fps. :slight_smile:

So, are you using Open GL through Rosetta2? Would be posible to use native Open GL? Thx.

Hey @unik!

I have it working natively ( not with Rosetta2 ) so I am assuming the OpenGL calls are native call too.



Is there somewhere we can see the compatibility of OFX through Rosetta2 (or natively!) for individual features?

You never know with new architectures…I’m interested in webcams and video playback specifically.

Thanks a lot, @theo!!
Good news that OpenGl still working on the M1 Big Sur. I will definitely go for the mbp m1.


See this post for a native build of 0.11.0:

1 Like