IMHO, Apple has screwed the pooch here. I worked a a computer company that insisted that executables linked on V1.0 would run on any subsequent OS version—but those days are long gone.
My suggestion for Ithinkiam is to look for a used machine on the refurbished market to hold your daughter over until the Apple chaos subsides. I have an olde Mac that needs replacing but I’m holding off until things settle. I don’t want to be an OS tester.
I would be leery of getting an M1 machine at this stage anyway. It appears to me as an outside observer that Apple has not worked out how large memory systems (like those needed for Rhino) will be laid out architecturally and how graphic acceleration will finally function.
I am an architecture student with the M1 MacBook and I was wondering if the experimental version 8 of rhino is useable on my Mac. I’m not using mesh’s or anything but I do often use the rendered and shaded view options. Will the experimental version work for me in your opinion?
i would be interested either, but was not bold enough to buy an m1 yet, (by the way it always reminds me of bmw’ M1 Version of their cars)
my guess is that Metal which is now under development and is available in the WIP 8 Version could maybe make it work… just a wild guess and you would have to activate Metal with the command TestMetal as described here
Metal is the name of the graphics language Apple is moving to.
This has nothing to do with rendering materials whatsoever.
This thread is about Apple’s new Silicon M1 processor Macs that are not supported by Rhino.
Sorry for the confusion.
Has anyone seen any real Silicon Mac Architecture documentation (if anyone can remember the old VMS architecture handbooks)?
It appears to me that the Silicon Macs are memory constrained by current standards. The normal offering is 8GB while the largest I have seen is 16GB
At the same time they describe this as “unified memory” which appears to be shared by the GPU. My current machine is 32GB + 4GB on the GPU so this looks like a major step downward.
It appears from my reading that the memory is on the CPU chip, apparently an attempt to bypass the system bus that is the bottleneck of nearly every computer system.
I have seen no distinction between this unified memory and CPU cache. Would this unified memory be, in effect, a giant cache, or is it something distinct. My suspicion is the latter.
Be that the case, it seems possible they would run out of space for memory and need a separate class of memory that could sit out of the CPU. Rather than paging to secondary storage, the operating system might page between primary and secondary memory.
In any event, it is clearly Apple Silicon will be at the low end and it will be a while before we see such systems targeted towards the Rhino user.
Just came here to report that the curve drafting and layout tools have been working well with 7.6.2 + Big Sur 11.3 through 11.4 now. I am currently using a low spec m1 Mac mini and running a large 2D / 3D project with over 20 big layouts.
Cycles 3D performance is not great. I am no technical expert but I am assuming that OpenGL on Rosetta 2 has a large overhead. I mostly use Blender Cycles or Octane as they are performant on Apple Silicon. To be clear, I work with large models with many features, I think average 3D use would be fine. My coworker modelled a moderately complex thing on her M1 MacBook Air with no troubles. I can post a video on YouTube soon and link it here for any interested parties.
if you have a smaller window like you did for the intel version the performance increases significantly. for the m1 version you had the window reasonably bigger, also you zoomed in much closer to all the parts in the m1 version of that video, so i am not sure if that is a useful comparison.
Actually I’ve tried different modes to bring M1 to life with Rhino and all you mentioned about screen size and zooming sounds funny. Also I’ve running Rhino under the Parallels 17 in a Windows 11 virtual machine and have to say got the same performance results.