on 2022-08-05 10:00 AM - edited on 2024-01-24 10:42 AM by Noemi Balogh
In July 2022, Graphisoft released the Technology Preview version of Archicad 26 for Apple silicon. The package was made available for public testing.
The Apple silicon native version will allow Archicad 26 to take full benefit of the Apple silicon architecture, delivering faster performance while consuming less energy. This article will provide some comparisons between running Archicad 26 natively onApple silicon and running Archicad through Rosetta 2.
For more information about support for Apple silicon, please check out this article.
For the benchmark tests, we used a Macbook Air (Late 2020) equipped with:
Please note that this Macbook Air does not have a cooling fan. Under long heavy tasks, the chip will slow itself down as temperature rises (throttling behavior).
We performed the same sets of tests using two different Archicad versions on the same test machine mentioned above:
The test files included:
The graphs below show the performance improvements obtained during the tests, with results in the range of 28 to 55% increase in speed (baselines values from Archicad 26 3001 running through Rosetta 2)
Due to the nature of being a Technology Preview, please notice the following points:
We are constantly running new tests, so stay tuned for updates on this article. Tests on M1 Pro and M1 Max will come shortly.
If true, that would means that it's a VERY big performance boost.
the M1 was already about as fast as Intels Macs.
We are already very excited about the long-needed Apple Silicon version. Too bad there is almost 2 years delay for such an expensive software.
My question:
How does the performance look like with a Mac Studio Max and Ultra. Is it worth the almost double purchase price for an Ultra. Or is a Max enough? Can the new version even fully use so much RAM and CPU/GPU power?
IMHO, it depends of what you do with Archicad.
Archicad is a very CPU dependent software, but it doesn't scale that much with multithread.
Strong single core performance always seems to have a biggest performance impact than the number of cores (they got an impact, but less significant).
GPU is pretty much used only for rendering 3D pics.
If you (like us) don't use at all the Archicad render, GPU is not that important and the difference of performance between M1 Pro and Max is poor.
We got both Pro & Max and I can't see the difference pour our kind of projects.
I don't got an ultra to compare, and ultra does got a very big CPU boost (double the cores) but I don't think the difference will be that noticeable on 95% of projects especially with the price difference.
Ultra costing basically the 2 max, I would stick to the max and changing the computer sooner if needed.
I am not following the graph. It is supposed to be comparing the two chips, but it is comparing the three projects for differing tasks. What am I missing? Where are the bars for the the other computer (which ever one is not shown)?
The values of the bars in the graph represent % improvement of native Apple silicon over Rosetta.