Kurios Creative Media Logo

M1 vs M2 Max ZBrush & Blender Comparison Review

28 August 2023

Can the new Apple Silicon chips handle professional 3D sculpting work in Zbrush? What about using them for doing 3D print work?

Apple has unleashed the M2 Max, promising even more power and efficiency. In this review, we’ll be comparing it to the new M2 Max chip in the Macbook Pro to see how much of a difference there is, and whether or not it is worth the upgrade. So, stay tuned as we delve into real-time viewport performance, Blender cycles GPU rendering, ZBrush decimation speeds, and overall sculpting experience on these new Apple Silicon chips.

The Problem with YouTube  Gear Reviews

There’s been a ton of review videos already benchmarking the M2 chips. However, I’ve actually found that sometimes these YouTube tech reviewers get things wrong as they just run benchmarks on gear, but don’t actually use it long-term in professional production. 

Benchmarks have their place, but they’re different from actually using the gear for real-world work. Sometimes looking purely at benchmarks can be deceiving. So, I want to help other creative professionals with honest reviews on gear I actually personally use in real-world production. This is why my reviews won’t come out right as gear is released because it takes some time to use it and properly get a feel for it.

As creative professionals, we work hard for our money and want to make sure we’re spending it on the right tool for the job. As digital sculptors, we know that every stroke and detail matters. So, we need tools that will not get in the way of our creative flow. For me, the majority of my 3D work is in Zbrush, sculpting for 3D print, toy prototyping and designing collectibles. I do some rendering in Blender just for showing stills to clients or for my portfolio. So, I am not usually doing a lot of work which is heavily GPU intensive.

(Please note, all links in this article are affiliate links)

Why Choose a Mac over a PC?

We all know that the PC crushes it when it comes to GPU performance, especially with Nvidia’s monster video cards like the RTX 4000 series. There is simply no comparison, and as great as the Apple Silicon chips are, they aren’t going to compete with the raw performance of such beastly cards. So, if you’re looking for amazing GPU performance for gaming, real-time development work for video games, or just for spec bragging rights—then go buy or build a PC. However, the Apple Silicon Macs are by no means slouches in terms of graphics performance. They just aren’t made to do what the RTX GPUs are made for, and you’ve gotta understand that upfront. They’re different tools for different use cases.

However, if you’re reading this, you’re probably a Mac user or potential Mac user who is looking at the M2 chips for other reasons. I know that’s me. If I wanted to do crazy GPU-intensive stuff, I’d build a PC with an RTX setup. I used to be a PC-only guy for a long time while I worked in the 3D industry exclusively. I was the guy who said he’d never switch to a Mac. But I’ve since started doing more video, illustration and design work, which made me consider the Macs, especially with the switch to Apple Silicon. There are some things about the Mac ecosystem and the Macbook Pro that appeal to me other than just raw performance such as:

  • Portability and efficiency – you can actually do work on a Macbook Pro all day unplugged! Especially with video editing work and the built-in encoder/decoders, it makes doing video work on the go actually possible and a pleasant experience!
  • Quiet – I love how silent they are sitting on my desk. I do a lot of podcast recording, so having something that doesn’t sound like a jet taking off is important to me.
  • MacOS – I really like the feel of MacOS as well as many of its features. It just feels a lot more creator-friendly than Windows in my subjective opinion. I just feel more at home as a creative on a Mac with all of the little details and polishes they do in the OS. Plus the app store is awesome… there’s pretty much an app for everything.
  • Integration with my iPad Pro – using AirDrop or being able to move my mouse and keyboard seamlessly between my Macbook and iPad works like magic! It’s so cool to be able to sketch in ProCreate or sculpt a concept in Nomad 3D on my iPad and drag it over to continue on my Macbook. (By the way, being able to sculpt on an iPad from your couch is pretty cool… let me know if you’d like me to do a video on my workflow with that later) Those who are invested in the Apple ecosystem of products know just how smooth they make the integration between their products!
  • Aesthetics – let’s face it, as creatives, looks matter to us. And Apple has really sleek design that appeals to my design sensibilities
  • Build quality – everyone has to admit that Apple makes great laptops that look, feel and function great. Plus I’ve still yet to find another laptop trackpad that I can actually get work done on and not feel totally handicapped.

Anyways, all that to say is that there are a ton of other reasons why someone might choose to buy a Mac over a PC, and vice versa. At the end of the day, these computers are just tools. And the tool doesn’t make the artist – so let’s not get caught up in gear debates.

The purpose of this review is to help you determine whether or not the M2 Max would be a good choice for your use case or worth the upgrade if you already own the M1 Max by sharing some of my own experience.

Here are the specs of the computers I’ll be comparing:

  • M1 Max Mac Studio, unbinned version with 10-core CPU, 32-core GPU, 16-core Neural Engine, 32 GB of unified memory, and 1 TB SSD.
  • 16” MacBook Pro M2 Max with 12-core CPU, 38-core GPU, 16-core Neural Engine, 32 GB unified memory and 1 TB SSD.

So, both of them are the unbinned versions of the M1 and M2 Max chips with otherwise similar specs. The main difference is that one is a Studio desktop and the other is a laptop – which may have some implications on thermal performance. Just FYI, I quit all programs except OBS to record my screen and Notion to write some notes while running all my tests. So those programs may have been taking away a little bit of resources.

Blender Rendering

Let’s start off by comparing how the M1 and M2 Max chips do in rendering stills in Blender. For this, I decided to go with my THING model for consistency – since that’s what I used in my previous tests. It’s a game model that I built that’s normal mapped and textured at 4K resolution.

M1 Max rendering

With the M1 Max chip on the Mac Studio, we saw great viewport performance. It was able to rotate around the model and do all the tasks you would need to in the Blender viewport no problem. Similarly, on the M2 Max Macbook Pro, there was no slow-down in the viewport. Both handled it flawlessly. So, if you’re primarily doing similar character work to me, then you’re not going to have any problems with either of these. However, if you are working with heavier models, the M2 will perform slightly better than the M1 due to its higher efficiency and more CPU and GPU cores. But, it’s not going to be a crazy jump, in my opinion. (After all, the M2 really just seems to be an overclocked M1 with a few extra bells and whistles) So, I don’t think that that’s a strongg enough reason for upgrading if you already own an M1 Max.

What about rendering?

Doing a full body render of the model in 4K UHD, using Cycles with the GPU and samples set to 1024, the M1 Max took 1 minute 33 seconds on its first render. 1 minute and 29 seconds on the second, and 1 minute and 32 seconds on the third. So, it averaged about 1 minute and 32 seconds per frame. Not bad. But obviously, if you’re going to be doing a lot of rendering, you’d really benefit from an RTX 4000 series card that’s going to through a lot of muscle power at it. But, if you’re just doing light rendering or stuff for your personal reels, etc., then the M1 Max might be just fine – once you can control your heart’s temptations to envy. Remember, playing the comparison game is a joy killer!

M2 Max rendering

Loading up the same project on the M2 Max Macbook, it had the same great viewport performance and it averaged about 1 minute 25-30 seconds per frame. So, not really a huge jump, but there is a little extra power with the extra cores in the M2. I wouldn’t say it’s life-changing, but if you’re rendering a lot of frames, those few seconds can add up.

I will note though that at this point I was suspecting that the CPU and GPU were not maxing out for the renders. What was surprising to me, was that in some other render scenes, the M2 Max actually was slower than the M1 Max. I think that this may have been because of some performance throttling in the laptop version of the chip, but I can’t be sure since on those other tests I wasn’t as exact in setting up the scenes the same way on both machines.

However, for me, I’m pretty content with the results I got and if I end up needing more rendering horsepower for my professional work, I’ll invest in building a PC rig. After all, if you’re making money with this, it’s part of your business expenses to get the right tools since time is money. We all know that these Macs are not going to come close to a spec’d out PC in rendering. But for now, for just showing still renders of 3D sculpts for print (which is my main work currently), both the M1 and M2 were fine for me. That may be different for you, depending on the work you do.

ZBrush Performance

So, just how much can these Apple Silicon chips handle in ZBrush? What would happen if you pushed your Zbrush sculpts to the max with high-frequency details and poly counts?

I decided to find out. Unfortunately, I can’t show a lot of my professional work because of NDAs and I don’t have a lot of time for personal projects at the moment. So, I used some of my old personal work for these tests.

M1 Max sculpting

I loaded up my ZBrush sculpt of THE THING. Since I used it for doing my render tests in my last test with the M1, I figured it might be helpful for comparison. The model is only 29 million polygons in Zbrush – so nothing crazy. The head is about 5 million polygons, and sculpting on it is smooth as a hot knife through butter. But that shouldn’t surprise anyone – after all, even the most modest of computers should handle that.

So, I subdivided the head as high as it would go to 95 million polygons on a single SubTool mesh. (Zbrush limits you to 100 million polygons per SubTool) Then I tried sculpting on it with the Mac Studio M1 Max and it stuttered a bit and was a little laggy, but it ended up being workable if you needed to do something that crazy.

I then subdivided all the other SubTools to max out this model at 362.8 million polygons. I’m not sure when I’ll ever need to have a ZBrush model this high poly, but hey, why not! YOLO. As expected, it was very laggy, but it worked! I was able to drag out an alpha and sculpt. However, the rotation around the model was pretty slow and I wouldn’t recommend working like this.

This was a pretty respectable performance in my opinion, and it makes sense since ZBrush mainly uses your RAM and CPU – and the Mac Silicon chips with their insanely fast unified memory and great CPU performance handle it well. There’s not much advantage to having a beefy RTX card in ZBrush anyways. The real difference is going to be all RAM and CPU dependent.

M2 Max sculpting

I loaded the exact same model into the M2 Max Macbook Pro and did the same tests.

Obviously, sculpting on the 5 million polygon head was a breeze. Sculpting on the 95 million polygon head after subdividing it to its max was also very smooth on the M2 Max. Definitely felt better and smoother than the M1 Max.

However, once I started subdividing the other subtools to their max, I started seeing some differences. I subdivided the other subtools until I hit the same 362.8 million polygons in the scene as when we did the test with the M1 Max, but the performance seemed quite a bit more laggy on the M2 for some reason. This was when I started looking at the monitoring app I have and saw that it seemed like the CPU and GPU weren’t really being maxed out. ZBrush uses mainly CPU, so I was expecting the M2 Max to do at least as good or better than the M1 Max. After all, this was what all the YouTube tech reviewers were ranting about how the M2 chips were that much faster than the M1s. However, in my tests, this was not the case. It actually felt slightly worse, at least for the Macbook Pro version I had. So, I dont know if there’s some throttling going on in the laptops that doesn’t happen in the Mac Studio. But that might be something to note if that matters to you. I don’t have the money nor interest to go buy every version of the M2 in the Studio and Macbooks to do comparative testing to see if there actually is throttling (after all, I’ve got a business to run!), but this was just my experience so far.

In the 362.8 million poly scene, it was very laggy. You were able to sculpt, but not really rotate much. I would say it was not even workable. You’d have to hide your other subtools to rotate around the subtool you’re working on at an acceptable rate. However, remember, that these subtools are ridiculously subdivided, beyond what the average model would be that I work on professionally. This was just for the sake of testing. When the subtools were at more reasonable poly counts, the performance was fine.

I decided to keep going for the heck of it and subdivided all of the subtools to their max to max out the total polycount. I ended up with a scene with over 700 million polygons spread over 12 subtools. Amazingly, Zbrush didn’t crash. But also, it didn’t really rotate either. It was totally unusable. But also, I’d never create a model like this for production! However, if you isolated your subtools, even in the 700 million polygon scene, you could rotate and sculpt no problem – even on meshes above 80 million polygons. So, I guess you could do that as a workaround in a pinch.

M1 Max decimation

Just to make sure that it wasn’t just that THING model, I loaded up another model – this time a dragon sort of creature I’ve been working on.

This model is at 36.86 million polygons, with the main skin of the dragon sitting at 35.6 million polygons. I’ll probably break it up as I continue on the sculpt later on. But sculpting on this with the M1 Max was a piece of cake.

I subdivided the wings to 70 million polygons and sculpted on them with no problem also. The model was totalling 107 million polygons at this point.

I then decided to run some Decimation Master tests, since a lot of my professional work is doing sculpts for 3D printing for toys and action figures and you have to decimate your STL files to send to print. So, this is something where some increased performance would actually be helpful in my professional work.

I started by decimating the 35 million polygon body SubTool. The pre-processing took 6 minutes and 15 seconds. Not bad, but also not stellar. I then decimated it and it did it in 22 seconds – which is great. So, I’d say that for doing decimation for 3D print work, the M1 Max gives a respectable performance – but not top in class. Perhaps if you got the Ultra or more RAM it might give you some better results.

M2 Max Decimation

So, I ran the same tests on the M2 and as expected, it performed well also. This is a pretty easy sculpt to work on.

Running the decimation tests, here were my results and where things continued to get interesting…

When pre-processing the 35 million poly body of the dragon, the CPU was maxed out and GPU was sitting at around 33% usage – possibly because of OBS running in the background. Also, you heard the fans spin up – it wasn’t crazy loud like my PC laptop, but definitely audible. Those YouTube tech reviewers who boast that you can never hear the fans don’t know what they’re talking about or have never used it for actual production work. Because I’ve definitely heard the fans spin up on both the M1 Max Studio and the M2 Max Macbook Pro in my work. I will say though, to its credit, it’s definitely a lot less and quieter than any PC laptop I’ve used though. However, if you push the CPU or GPU either with a tough video render or CPU-intensive decimation, you’ll hear the fans spin up on the Macbook Pro. So, not totally silent, but also a lot quieter and less frequent than PCs. The Mac Studio though, did sometimes spin the fans up, but slightly less in my experience than with the Macbook Pro. So, I definitely think there’s some cooling differences between the two.

Anyways, back to the test. The pre-processing for the decimation took 6 minutes and 58 seconds. So, actually a little slower than the M1 Max Mac Studio! This was surprising to me, but again pointed to the fact that perhaps the Apple Silicon chips in the laptops are not 100% exactly the same performance as in the Studio. I have to think that thermals must play a role here. So, if I were to guess, I’d guess that the 14” Macbook Pro would perform worse than the 16” because of thermal performance. It just goes to show, there are always trade-offs. If you want portability, you have to trade off something—either performance or battery life. And that’s true for Macs or PCs.

Back to the test, the decimation took 18 seconds. So while the preprocessing was slightly slower, the decimation was definitely faster than the M1 Max with those 2 extra cores. This shows me that it’s moreso the sustained stress that causes the M2 in the laptop to throttle down, but for burst performance, it is faster than the M1. This may not happen in the Studio version of the M2 – but I can’t confirm this as I don’t own one.

Rotating around the 7 million polygon decimated model is laggy – but in production, I’d probably decimate it even further to deliver to a client. When I drop to 2.8 million polygons decimated, it’s fine. And at 1.4 million – which is closer to what I’d probably deliver for 3D print, there’s barely a noticeable lag. So, for my work, this is fine.

Remember that the tests I’m running here are pushing well beyond what I’d normally do for production work for clients. I almost never will have a single mesh that’s more than 40 million polygons, most of my ZBrush projects are under 200 million polygons, even for complex models. So, while it’s nice to boast about specs and how many polygons your computer can push, realistically – how many do you actually need for your work?

Concluding Thoughts and Important Notes

In summary, both the M1 and M2 Max chips can handle quite a bit of ZBrush work and could be used for professional work. I’ve been using them with no problem!

The main reason for me selling my Mac Studio and upgrading to the M2 Max Macbook Pro was that I needed something portable since I also do video work, and wanted a laptop to have with me on shoots. Also, being able to be mobile with my work and realistically do work on a plane or at a coffee shop was a huge appeal. As we know, most PC laptops with beefy specs are going to be faster, but really don’t have great battery life. I’m still blown away by the Macbook’s battery life – I easily can get a full day of work on one charge… and being able to sculpt in Zbrush or edit 4K video on the go is pretty great!

Some important notes and tips though:

I did find that sometimes Zbrush has to pack or dump the memory when working on extremely large files. I was working on a sculpt of a mech with lots of different parts for a client, and Zbrush had to delete the undo history because it ran out of RAM. So, if that’s a concern for you – you might want to consider upgrading to the 64GB of unified memory. However, this is not to say that it’s unusable. It worked just fine for me and I completed that project without issues. Once you’re saving iteratively (as you should be), there’s not much issue there. If I had the extra budget, I would have gone with the extra RAM. But for the vast majority of my work, I hardly run into problems.

Connected with this is the issue of using SWAP Memory. Sometimes as the computer runs out of available RAM, it will write to the swap memory – that is, the solid-state drive. It gives you a little pop-up warning when this happens. However, this also hasn’t really affected my work even when it happens. The SSDs in the Mac are super fast! But, again, if you’d like to totally avoid this, then upgrade the RAM. It’s just gonna cost a pretty penny since you pay a premium with Apple for more memory and there’s no way to upgrade it yourself after the fact.

I’d say that for the majority of people 32 GB will be fine. For certain professionals who are working on extremely heavy projects, it will be worth buying the 64 GB version or 128 GB. For me, it wasn’t worth it and I’ll probably end up building a PC rig if I get to that point anyways.

You should, however, upgrade the SSD to at least 1 TB in my opinion. 512 GB will fill up way too fast with apps and stuff you’ll need on your internal drive. Beyond that, I’d say buy a couple of external Thunderbolt SSD drives to save some money on storage space. I’ve been working off of these for 3D and video work with no problems and they are way cheaper per GB than Apple’s internal storage pricing. I’ve used both the SanDisk Extreme 2 TB and Samsung T7 2TB drives, and while I know there are some minor differences in read and write speeds, I honestly can’t tell the difference.

If you want to get the absolute most out of the M2 Max chip, it seems like you should get the Mac Studio – since the M1 version didn’t throttle (I’d assume that the M2 version also wouldn’t throttle because of its superior thermal performance). However, even with the slightly throttled performance on the Macbook Pro, for me, it hasn’t been that much of an issue using it for professional work. The other pros that come with it – like portability and battery life – make up for it for me – and at the end of the day, it’s a really great machine that’s capable of quite a lot. So, if portability is important to you, the Macbook is a great choice, but if you need the most performance, the Studio is the better choice because I don’t think the performance is exactly identical in real-world use cases.

I’ve been using the M2 Max Macbook for over 4 months now on professional client work in both 3D sculpting and video editing and I definitely don’t regret my choice as it has proven to be the right tool for me for the work I currently do. There may come a day when I’ll need to upgrade to something else, but that’s how business in the tech industry is—so don’t get hung up on it. It may be that it will be the right tool for you, or not – that depends on your use case and what features are important to you. At the end of the day, these are just tools. Get the one that works and fits your budget, get off YouTube and the comment debates, and go create something awesome!

error: This content is protected.
Share This
Seraphinite AcceleratorOptimized by Seraphinite Accelerator
Turns on site high speed to be attractive for people and search engines.