OTOY Forums

Strange result. Apple claims that M2 Ultra is many times superior to M1 Ultra

Thanks Jose for sharing!

Damn, I was expecting at least a 30% boost

Here is the updated graph, I have included a Nvidia RTX 2070 Super result: ciao,
Beppe

Hi all,

I've been following this forum for more than 2 years and seeing the latest improvements and benchmarks, which are impressive, decided to do some of my own test runs. I'm a professional 3D designer and use C4D + Octane daily. As many on this forum, I'm patiently waiting for when the Live Viewer performance on a Mac will reach a point of being comfortable enough to use in daily work (comparable to RTX 3060 as a minimum).

I tested the VolumeDenoiser scene on 3 machines and here are the results:

• - RTX 3060 (Win 10) - 0:14
  - M2 MacBook Air (8gb) - 1:36
  - M1 iMac (16GB) - 2:00

I used Octane 2022.1.1. version both on Mac and Windows. Hope this helps!

SammyDH74 wrote:Hi -

All I want to do is make my GPU available for the network. So, if we have a new Ipad Pro with an M2 does this suffice to make the GPU available as a "node"? Please advise.

Thanks!

Can someone please answer this? Thanks.

SammyDH74 wrote:

SammyDH74 wrote:Hi -

All I want to do is make my GPU available for the network. So, if we have a new Ipad Pro with an M2 does this suffice to make the GPU available as a "node"? Please advise.

Thanks!

Can someone please answer this? Thanks.

I can't speak for Otoy but in my experience, as GPUs have gotten faster and faster, network speed has become more of a bottleneck. When we upgraded our in-house mini render farm to 4090s, it became clear that it was time to upgrade our gigabit network to 10 gigabit.

From what I can tell, the newest Mac CPU/GPUs are roughly 2x as fast as the 1080 ti cards we upgraded from. I suspect that wireless speeds may not be fast enough to make using an iPad really worth while, at least not for basic lookdev and test renders, but maybe for big, slow rendering frames. It's certainly worth testing, but I wouldn't get my hopes up.

I'm totally talking out of my ass, though, so take it with a teaspoon of salt.

This chart is not correct.
mac studio everything is very sad
My test:
6900XT Mac pro 9 seconds
3070+ rtf on 8 seconds

I want to buy Mac studio 72 core. but I can not understand the performance. Will it be faster than my 6900xt.

Perhaps we can see how fast the M2 Ultra renders the Chess scene.
This was the previous Octane X benchmark test that didnt use noise reduction. I dont use the noise reduction as its so slow on intel.

I have x2 6800xt as EPGUs. Its pretty fast and I use it for commercial work. Unsure if the upgrade to the M2 ultra is worth it.

@spybreak @otoy

Is there any other more complicated test? It feels mostly a denoiser and volume performance, and would be great to see more complex scene

Hi everyone!

Just for comparison - I ran the test on my ASUS Zephyrus M16 (RTX 4080 Laptop):
Silent profile (fans off) on battery: 12 sec
Performance profile on battery: 10 sec
Silent profile plugged in: 10 sec
Performance profile plugged in: 8 sec
Turbo profile plugged in: 7 sec

But I think that such a short benchmark is a bit pointless - why is there still no OctaneBench X?

Are there no AMD updates planned for the future?