Hi everyone,

We've added some new binaries for this project to allow macs to run.
Both x86 and ARM64 based macs should run just fine.

The linux and windows binaries remain at v1.05, there'd be nothing new for them if we did bump them to 1.06.
Macos is on 1.06 because I uncovered a bug in my deployment of the macos binaries and had to use a version bump to replace the old version.

Anyway, something interesting I've discovered is that M1 (and really, all apple silicon so far) LOVES these tasks.

On a sample mini task, I got 120 seconds on my 12700k, and 61 seconds on my M1 macbook air.
On the same sample mini task, one of our project developers tried it on their M4 Pro and got 43 seconds.
Of course, that's not perfectly reflective of runtimes, but I expect to see apple silicon based machines top the charts for this app based on these preliminary results.

Happy crunching, everyone!

They can try to take down Genoa all they want. LOL

M1 almost beats the fastest task time (which was achieved by a 7950X3D) for config 53.
(I was using it as a baseline)

M4 Pro beats that fastest task time :)

It might be the fastest task times, but it's going to take a lot to take the top computer off the top spot on the charts. I doubt the M4 Pro (including the Max) will have enough cores to do that.

What about Apple Silicon GPU cores? Or is that not possible for you to support due to hardware limitations or something?

Also, despite me getting 10 concurrent tasks on my base M4 Mac mini, 1 for each CPU core, and 98-99% CPU utilization, temps, wattage, and clock speed aren't as high as they are for other projects, that also have native Apple Silicon support, and run at max utilization.

Don't know why this is, maybe the WU's just aren't as demanding, despite utilization the CPU fully, but maybe there's some room for optimization, for the Apple Silicon chips in general, or just the M4 chips?