Apple, in the past was always saying they are into content creators,
while a typical Windows user is just a poor content consumer.
(Quicktime, Adobe, Macromedia, …)
But that changed when Apple went started to sell or even produce content
and services to be consumed. Basically with the release of iOS.
And they saw large markets.
At least beginning with 2010. That was when they lost interest in (Pro)
content creators. They went to casual content creators instead,
Blogs, Selfies, youtube influencers, … boutique and fashion products
where the value and margin is measured or given by emotions and attention.
At iPod and Apple Soft and Hardware times this was going hand in hand.
Much better usability for users, high quality standards, basically realistic
and realistic prices made products selling well and gaining hight revenue.
But with 2013 Mac Pro the produced a Final Cut Video Station only hardware.
Scared Video Editors with a half baked Final Cut X and such things.
Apple may have already have thought about A-SoCs in Macs t that time.
But neglect whole professions supply for a decade isn’t that empathic.
Apple made us (the last few) hope and still not leave the ship like most
reasonable creators, by an also financially and power wise not very
attractive interims iMac Pro (some like it !) and a caricature of a Mac Pro later.
If you don’t need it and make business you won’t earn it.
And now Apple had a famous promising M1 Start, basically more potent than
latest Intel Macs from the start, but more than two years later stagnating at
a Pace, to be used by Intel at its worst competitive free years idling from
(2011 ? - 2018 ?)
15% of increase in 1.5-2 years for M2 chips by adding more cores, more
frequency and more power consumption feels a little poor for me.
After M1 Max/ultra not even scaling well.
I do not believe that Apple should come up with any M2 Pro/Max below
a 5 nm process and major speed improvements.
Apple’s M1 Max GPU is very strong in raster performance. I know RT is the new hotness but for an AR headset I don’t see the immediate benefit. It would make even more sense for them to have something they could use not just RT but other calculations as well if necessary. Nvidia and AMD can afford to add silicon to their cards just to accelerate RT tasks, a headset with a given power envelope not so much.
I feel that the Mac Studio is a fantastic machine for 3d in Blender. Highly performant viewport and animation speeds. It’s great EXCEPT for rendering. Compared to even a lower end Geforce 3060, my M1 Max chip gets lapped by it. However, if you need raw processing power, it’s solved with just a cheap pc build with a big-ass card sitting in the other room. It has additional advantages of being able to burn frames all day with without slowing down your ability to keep working, so I’m really not going to lose much sleep over not being able plug an eGPU in.
However, we should also not forget that AS optimization has barely begun. Once it’s fairly well optimized the M1 Max has the potential to get close to my Geforce 3060 for rendering, which would be VERY impressive IMHO. The 3060 will still win, due to those raytracing cores, but again, that kind of rendering speed in such a small quiet package would be simply awesome. I can let the vacuum cleaner in the other room handle the bulk of the rendering workload.
I’m betting that a native metal EEVEE will perform extremely well also.
Additionally, a typical freelance workflow is more than just blender. For motion graphics, editing, image editing, painting, etc, its just awesome, and using MacOS for all of that is a terrific workflow for me.
Whenever I read ‘Hackintosh’ I always think of above-Linux-level maintenance hassle for tech wizards, as there’s no Apple support for such a custom hardware configuration. But maybe that’s not the right impression?
Well written. This is the reason I’m keeping a close eye on Apple Silicon developments, both on the hardware side (possible new Mac Studio configurations) and the software side (Metal optimization by Apple devs, native Metal support inside tools, etc.). Until then I will have to keep listening to my Windows vacuum cleaner when rendering.
Yes sounds like a bit of tinkering, but less the Linux I think. I would perhaps have tried it but Amd CPU support on Hackintosch is not amazing and a lot of features are broken.
Which is normal as Apple never had AMD CPUs
I put the opposite in my PC of what you should for a Hackintosch
So basically the way it works is relatively straightforward. You assemble the machine from a list of “approved” hardware (although surprisingly AMD CPU’s are also configurable even though Apple has never used them).
The most “Linux-like” thing is configuring the Open Core doc which is basically a text document with a series of “on/off” options. It looks daunting at first, but there are plenty of step-by-step guides that make it a fairly easy to follow thing to do. In addition, there are plenty of people who will share their configuration files (assuming your hardware is the same).
Beyond that, the machine runs like a regular Mac, all the updates work automatically, there are no special day-to-day requirements.
Having said that, I would love to go to a smaller/simpler Mac Studio setup, but only if they re-enable eGPU. Without it, the render penalty is simply too high. I don’t do much animation, so for me it’s all about looking at full renders to evaluate lighting, materials, and scene composition. In addition since I primarily work in film, I need to render out at times thousands of frames, so that render time is really critical.
The idea of sending out the render to a second Nvidia machine is also not that appealing. I actually do have a second computer set up for that, and I hardly ever use it because of the hassle of keeping the two machines perfectly in sync with files/materials/add-ons/etc.
Don’t forget that Mac Pro 2009-2012 is in fact Hackintosh these days. You have to pass all these Open Core steps to force it to work with the newest macOS versions.
One more thing regarding fast GPU. It is not only about final rendering. It is also about instant rendered scene preview BEFORE final rendering. For some, this is absolutely essential.
“The idea of sending out the render to a second Nvidia machine is also not that appealing. I actually do have a second computer set up for that, and I hardly ever use it because of the hassle of keeping the two machines perfectly in sync with files/materials/add-ons/etc.”
Keeping the two in sync is simply a matter of your machines both connected to a shared network location. The only hassle I deal with is just the upkeep of 2 machines vs one, honestly. But if you render a lot, then you likely already have multiple machines to help with the workload. Rendering an animation project on a deadline and trying to only use one machine to do it all is… not the best choice if you can help it.
That requires NAS which I haven’t looked into yet. I might in the future…but at the moment my render speeds are so fast that I hardly ever need to offload to another machine.
Gotcha. Curious why you would need a NAS unless you are editing 4k videos in realtime or something. Are you needing to move gigabytes of textures or something? Serious question. My workloads are generally pretty light so just wondering what of your 3d workloads require such high transfer speeds.
I have an external Data SSD via Thunderbolt 2 Enclosure
(Since Trash Can times) on my M1 Mini.
This is a Shared Drive for all my Macs and also for the PC.
Could be an internal drive or even just a partition of it though.
Yes.
During my Mac Pro 2.1 times I had basically only a single machine with
which I was wish less and happy for 3-5 years.
OK, there came a 17" MBP, not much used, mainly as backup solution.
But that was no problem as there was no problem keeping Macs up to date.
Before I always had 1-3 PCs running and cared more about updating and
supporting my machines than working hours with them.
This was a great time.
Currently I have my 4 Macs + 1 PC accessible and still working
Cheese Grater 2.1 2007, Mac Mini Server 2009, Trash Can 2013 and,
always on, my M1 Mini.
Just for memories, only the old Mini Server with its old OSX, to drive my old
32bit Scanners and Printers.
Yeah, that’s exactly why having a separate render-PC doesn’t appeal to me too. Even running Bootcamp still requires maintaining a Windows installation next to macOS, and I struggle with switching between the two OSes. My wife is still working with Mac, and whenever I sit behind her machine I constantly make errors like Command + Z versus Control + Z, and many other differences between macOS and Windows. Besides all that, I would love to completely return to macOS and leave Windows, preferably forever.
Having to maintain 2 machines is annoying in itself.
While having to maintain a PC is even another non-Mac level annoyance …
My Ryzen/RX PC is far more capable than my M1 Mini.
But that lead to do my whole project work on the PC for “simplicity” instead
of switching over for rendering.
And to be honest, each time I test Software, or certain SW features not available
on Mac or Apple ARM on the PC, I get annoyed again by how much worse and
unnecessarily complicated or simply bad Windows is as an OS.
I think Windows 11 finally is eye candy, but looking at daily workflows,
Finder vs Explorer, manual Software Settings Editing, SW un/installation,
Spaces vs Virtual Desktops, …
I could not work with a PC.
(At least as long as I have access to a Mac comparison)
I would rather go with my Linux Elementary OS.
(When do they finally get a rolling release !)
