Perfect, I'm curious about maya. Thats what I do most of my time so wonder how good the performance is there. 
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
3D Rendering on Apple Silicon, CPU&GPU
- Thread starter vel0city
- Start date
- Sort by reaction score
I am curious about thermals and fan noise during long sessions. I suspicious about using a Laptop.
I have been used M1 Ultra last 18 months and honestly there is no fan noise or heat at all.
Since last 6 months we have experienced a few issues with 16 inch M1 Max and a few weeks ago switched it to M2 Max Studio. Issues has gone and it is ok for now.
I am going to decide to buy or not for daily 3D/Graphic/Video Compositing usage after your experience about these subjects.
Also looking forward to user experiences rather than endless benchmarks. It also includes noise levels of the computer, but also performance of types of scenes, materials etc. Therefore I use my own benchmark as I work with lots of reflective and transparent objects in different combinations.
Honestly
I have gone to Apple store two days ago for buying an M3 Max Macbook Pro and I loved it at first moment.
Looks great, feels great.
Than I realized a bunch of thing.
I do not feel enough safe for long term usage about M3 Max Macbook Pro. Of course this is only my perspective, it might be great for someone else.
I believe rendering faster or already getting higher scores at benchmarks will not transform a laptop to a desktop contender.
At the end of the day it is a Laptop and I had to pay it more than 4.000 USD with taxes ext. but I need more ports for my gadgets, monitors, SSD's + I had to buy hub for more TB/USBC ports.
I am not sure for ZBrush/C4D/Texturing workflow, there would be issues...
If there is not 😁😁😁
Very much this. Any details about the ZBrush>C4D>Redshift pipeline, day-to-day usability, things like the time to first pixel in Redhshift, displacement maps from Zbrush to C4D/Redshift, particle sim calculations, X-Particles viewport speed, this is the real stuff that benchmarks don't tell us.
I observed common slowdowns on M1 Ultra triggered by high poly counts/large memory simulations/Large tex files.
Redshift is not ready for M3 HW RT for now it use raw power of GPU cores, first sights are really promising but there is no eta for full AS Hardware RT.
I do not use Blender much and it will not be a purchase reason for me.
Here it says it already does: https://www.maxon.net/en/article/ma...imized-for-mac-lineup-with-m3-family-of-chips
It is optimized only for Cinebench 2024 not for Redshift's current state. They will implement M3 HW RT soon I supposed.
I understand you completely. I said it before, the line between laptop and desktop is very small with Apples MX chips but to choose a laptop for rending tasks??? It feels odd. My 7 sec animations are rendered out in 1-2h in absolute silence on a M1 Pro. That is even odder but what a bliss. However, I do too little render to be used as a guidance for other but I also asked myself the same question- can MX laptop be used as a heavy duty compute machine? There is a minor increase in the temperature and no noticeable fan noise during render. Let the engineers answer this question.
We have experienced a issues with 16 inch M1 Max ( Davinci Resolve/Compositing machine ) which are TB Ports are overheated during time, started to not work as it has to be, fan noise is getting higher and higher day to day, Screen outputs started to blinked random again and again ext.
After an hour a M3 Max 16 inch could effected by performance but I am not sure about internals for long term.
Anyway M3 16 could be ok and if it is more than enough for me...
The new M2 I bought in September I kept in clam shell mode the whole time. I did my initial tests and it was fine. Yes doing a long render, for my workflow, (1-5 hours) the fans came on but were nowhere near like my old Intel MacBook Pro was. For everything else they never turned on. I suspect the M3 will be similar but much faster.
I see you are concerned about zbrush. What sort of models do you work with? All my creature models ran like butter with the initial M2 tests. I can certainly test a bust using HD Geometry and project some XYZ alphas with the M3. Pushing a billion polygons and I suspect it will run perfectly fine. Again, not sure what you do with zbrush.
Houdini grooming renders were exceptionally fast with my M2 and using Karma. Look toward to testing XPU with the new H20.
Thanks for your feedback
ZBrush is my oldest and most comfortable modeling/texturing software, I use it for anything I need on 3D
Currently M1 Ultra can handle over 40 million polygons for a single mesh. I do not have any idea how will it becomes for all subtools or HD Goemetry side. My concern is UV Unwraping/Texture Painting/Masking Tools all of them relies on cpu count/power. Maybe M1 Ultra not fast as M3 Max but has got 4 more cores.I render out scenes from a PC, I am not using M1 Ultra as a render device.
For Redshift Lookdev and shading process it is ok now for working long hours on M1 Ultra. I don not use Houdini, I am an old manfor start to learning it
I am asking all these quesntions because of I can not give a decide clearly for buying an M3 Max Laptop or waiting for M3 Ultra. I can not afford both of these machines at same time, I had to choose one of them
I feel like M3 Studio variants will arrive than expected timeline but on the other hand M3 Max is tempting right now hard to choose one of them...Apple has xmas extended return period so buy now and test it till first week of january. By then you will know exactly what you need. Also, in january you get full refund and then you get the machine based on your experience. Win win
Thanks for your feedback
ZBrush is my oldest and most comfortable modeling/texturing software, I use it for anything I need on 3D
I render out scenes from a PC, I am not using M1 Ultra as a render device.
For Redshift Lookdev and shading process it is ok now for working long hours on M1 Ultra. I don not use Houdini, I am an old manfor start to learning it
I am asking all these quesntions because of I can not give a decide clearly for buying an M3 Max Laptop or waiting for M3 Ultra. I can not afford both of these machines at same time, I had to choose one of them
good idea
I am going to check it, but it can be done only selected countries I guess. Even in europe it could be quite complicated.I've done it in Europe many times but yeah, its true that its only for some countries so have a look. Otherwise its the usual 14 days.
good idea
I understand. No problem at all. I’ll keep you posted. To be honest with you, I was in a similar situation. When I bought my M2, I was intending to then buy the M3 Ultra Studio, loaded. Then the new MacBooks dropped immediately after and I was like wtf. So now, after swapping machines, this will cover me for my needs. I won’t be buying a studio. The M2 Max was near perfect for what I was doing, so I am fairly confident that the extra resources and raytracing will really put this machine in a sweet spot for me. My rendering is not for animation, just still frames and perhaps turn tables. So yea, that is certainly different if I was rendering motion graphics or actual production animation. I can say this though, one of my personal projects was a likeness of Gerard Butler from 300. It’s pushing 100 udims in Maya, all 4k, with some 8K vface textures for the portrait, fully groomed with fuzz on the cape. Typically on my old machine it would take 10 minutes to reach first pixel. On the M2 Max it was a fraction of that, like under 3 minutes if I remember correctly, and the render was faster. This was with Arnold in Maya. No gpu rendering. So again, for what I’m doing, I think I’ve got desktop class performance, and then some, in a notebook. If I was rendering animated shorts on the regular, then yes, probably I would opt for a Studio. For pure asset creation, though, I suspect these machines are beasts. All the reviews I see online don’t help. People running bench marks of a blender scene. That’s the most I can see for 3D. But it’s pretty much useless for me when comparing the sort of 3D work I do. Omg this chip render the classroom 1 minute faster! You know as a professional that there is so much more in our work where we can save time. Oh yeah, I forgot to mention, the groom on that cape has like a half a million to a million fibres, I can spin the viewport much smoother than my older workstation. It’s stuff like that, that adds up in the long run.Thanks for your feedback
ZBrush is my oldest and most comfortable modeling/texturing software, I use it for anything I need on 3D
I render out scenes from a PC, I am not using M1 Ultra as a render device.
For Redshift Lookdev and shading process it is ok now for working long hours on M1 Ultra. I don not use Houdini, I am an old manfor start to learning it
I am asking all these quesntions because of I can not give a decide clearly for buying an M3 Max Laptop or waiting for M3 Ultra. I can not afford both of these machines at same time, I had to choose one of them
Oh wow! Thanks for this. I thought we only had 15 days!
yes 14 days period is valid, also mention about the date 6 january 2024 but not clear which products could be on this trade. Had talk with sales tomorrow...
cool I can not imagine 32 core M3 Ultra on Arnold, wish to work that way. It will be great for me using Arnold again. I have use Arnold for a long time
even use it quite much at first release before Autodesk. It was an extension for Maya shader system as I remember.I have started to work C4D with 2xVega eGPU's couple of years before and during Pandemic with Prorender, than shift to Octane and Redshift. I always prefer Arnold for any project or at least V-Ray but both of them have longer single frame render times for animation scenes.
Most of the reviews targeting as ads. Only couple of them useful.
Real Life workflows are quite valuable and uniq for each user, it is very hard to find real information gems in this pile of videos.
I do not like GPU rendering at all but this is only option for animated scenes for a while. Every frame need to be rendered out under 30-40 seconds, otherwise whole workflow is useless.
I've upgraded from my 2019 Mac Pro to an M2 Ultra, with the intention of upgrading to the M3 Ultra when it comes out. Amazingly, the M2 76-core is as fast (if not faster) in Blender than my Mac Pro and W6800X Duo. Since Apple won't give any more love to its Intel machines – and Blender, blaming Apple's lacklustre drivers, is aiming to drop support for AMD GPUs in the macOS version – I decided it was best to cut my (considerable) losses and move to Apple Silicon. However, one key upside, apart from slightly faster rendering, is that with the unified memory, I can create scenes with millions of instances that would utterly choke my dual GPUs. A pair of M3 Max chips should be very impressive.
How is the current version of C4DtoA? Have you used the Apple Silicon native version?
Yes I have been used it for some test, it is fast and reliable. Some scenes are faster than Redshift CPU and very rare but but even faster than Redshift however this is not a general rule you can accept for all. I guess depends on your workflow. It will be great on upcoming M3 Ultra with 32 cores maybe higher cores counts with M3 Mac Pro. Finally Arnold lovers could use it on Macs. 50 USD per mount or 400 USD for a year
Reasonable price V-Ray has got a little bit higher price tag.
Yes I have been used it for some test, it is fast and reliable. Some scenes are faster than Redshift CPU and very rare but but even faster than Redshift however this is not a general rule you can accept for all. I guess depends on your workflow. It will be great on upcoming M3 Ultra with 32 cores maybe higher cores counts with M3 Mac Pro. Finally Arnold lovers could use it on Macs. 50 USD per mount or 400 USD for a year
Thanks for that. Arnold is by far my favourite engine so this is great to hear. I just don't find Redshift's aesthetic pleasing or beautiful, Arnold's look appeals to me a lot more. I'll have a look at the new version of C4DtoA before I resub to Redshift. Arnold's Toon shader looks incredibly impressive, an area where Redshift is frustratingly lagging.
Agreed with you by all these subjects. You can download Arnold Trail from Autodesk with an account and try it first place. Lower CPU counts could be an issue, I never really on Arnolds GPU render actually it can be use only with NVidia GPU and available on PCs. Before subscribe to Arnold you have chance to compare it with Redshift.
I do lighter, more technical work, mainly in Houdini and have been using a m1max for the last years as my primary machine. It has been rough in many ways but also enough. When doing any serious work, I am usually connected to a display and have a lot of gadgets connected so I have been waiting for “the serious Mac desktop” for a while. When doing any sim or render that takes more than 5 mins, fans spin up. Not loudly, but enough to stress me about what happens with dust buildup etc. also, the battery gets worse by time so in the end it will become a desktop anyway.
The m3max uses a lot more power it seems, so fans will have to be working harder. The “lust” for this machine is high, but I will wait for the ultra. That will be a machine that can stay with me for many years. A m3max specced out enough is in the 7000$ range, way to much for a biannual investment.
The m3max uses a lot more power it seems, so fans will have to be working harder. The “lust” for this machine is high, but I will wait for the ultra. That will be a machine that can stay with me for many years. A m3max specced out enough is in the 7000$ range, way to much for a biannual investment.