Topaz Video Enhance AI benchmarks on Apple website!

I was looking at the Mac Studio M2 Max/Ultra specs website and I noticed that it has Topaz Video Enhance benchmarks under video processing. I thought that was interesting.

Will be interesting to see real life benchmarks from users.

2 Likes

As I suspected, the small print states the tests were for upscaling HD sources. Apple should be embarrassed, if not ashamed, at the M1 / M2 performance when upscaling SD sources. I would very much like to hear some technical, but understandable, explanation from Topaz why upscaling SD on Apple Silicon is so inefficient.

From the Topaz benchmarks, all other systems (Intel Macs and PCs) show almost exactly proportional performance between SD and HD upscaling.

Thanks.

Andy

How much money Topaz Labs LLC must have paid to Apple to appear on a measly performance chart that few people will ever see. :joy:

2 Likes

For those of you who are interested in sharing and reading other user’s benchmarks, check out this
thread.

@sergiohzph We do not pay Apple nor do we ask them to create or share benchmarking. Apple posts these types of articles based on what apps their user pool is using and are excited about.

3 Likes

It looks like you didn’t even wait to find out if the issue was Apple’s fault before declaring that Apple should be “embarrassed, if not ashamed.”

Should Intel have been embarrassed and/or ashamed when the early ports of Photoshop from 68K Macs to x86 PCs performed worse on the Intel chips? Turns out that really wasn’t Intel’s fault at all; later versions of Photoshop addressed the x86 performance issues. It’s not shocking when a CPU-intensive application that has been developed and optimized under one family of CPUs doesn’t immediately perform as well when ported to a new architecture of CPU.

I didn’t say Apple was solely to blame. I’m sure there are things that both Apple and Topaz could do to vastly improve SD upscaling performance on Apple Silicon.

Andy

Until I understand the root cause(s) of the problem, I don’t know whether Apple bears any, some, most, or all of the blame. Maybe there is something that Apple can do and maybe there is not.

Having wrongly blamed hardware for problems that turned out to be solely my fault as a young software developer, I tend to be cautious now.