Mac Studio Review: Maximum Macintosh, Redefined
There have always been 3 camps of Mac users — consumers who wanted a small all-in-one with decent specs and an “it just works” philosophy, pros who wanted a completely modular and large system that would handle anything they threw at it, and then the 5%ers — dubbed “the xMac crowd.” The xMac crowd wanted a computer with all the benefits of the consumer Macs, but with power that they could use to do their jobs. xMac lovers wanted a computer made just for them since the early 90s. But at Apple’s Peek Performance event, Apple finally pulled off the computer that this small group of Mac users had wanted for decades. Enter Mac Studio — Mac Studio takes all the power and capability of Mac Pro and puts it into a small factor aluminum cube that sits on your desk. It’s peak function-over-form, yet transcends the boundaries of what’s possible. Mac Studio does good on every single wish of the xMac crowd; it’s as powerful as the modular Mac Pro (or PowerMac at the time), it’s as small as the failed trash-can Mac Pro, it’s not attached to a display or speakers or camera, and it’s affordable. Mac Studio is the computer of the content creator, the xMac lover, and the enthusiast. It’s the computer that will fuel creative computing for the rest of the Mac’s life. So as a content creator who edits photos and videos daily, I feel like I should review Mac Studio from a different perspective — not as a computer, but as a tool. And more importantly, show the world what’s possible with this small, inconspicuous cube. Buckle up, because this Mac Studio review is unlike one you’ve ever seen before.
Apple has always been a company obsessed with design — and this obsession hasn’t changed with Mac Studio. Many people have criticized the design of the Mac Studio for being too simple — and to that I say… that’s kind of the point. It’s not extravagant like the Mac Pro, yet it looks appropriate for the amount of power it holds. Mac Studio is intended to sit on the right side of your desk. It doesn’t take up too much room, yet makes the desk feel more filled up. It’s not distracting when it’s on your desk — instead blending into your peripheral vision. But the main strong point of this design is the ports. Mac Studio is chock-full of ports; 6 USB-C ports, 2 USB-A ports, 1 HDMI port, an SD card slot, and a headphone jack. That might not seem like a lot on paper, but it’s more than enough for pretty much everything. 1 display, keyboard, mouse, backup drive, webcam, speakers, and audio interface would all be able to connect to the computer with ports to spare on the front. I find myself plugging in long-term devices like my backup disk for Time Machine, display, wired keyboard, and lightning cable in the back of the computer because they’re a bit hard to access while using the front USB-C ports for quick peripherals. I do have one complaint though — I wish there were more USB-A ports. I don’t know about most, but I’m not “pro” enough to have completely moved to Thunderbolt 4. I frequently want to plug in another USB drive, charger, and more into my computer, but I’ve already used up the 2 USB-A ports at the back for my backup drive and keyboard. I realize that this is somewhat my fault since I haven’t upgraded my cables to USB-C, but regardless, I would have wished for an extra USB-A port at the front at the bare minimum. Also, the HDMI port on the back of the computer is not HDMI 2.1 — just 2.0. I let this slide in my MacBook Pro review because the HDMI port was mostly intended for projectors and TVs (though you can make the argument that those don’t exist thanks to the pandemic), but this is unacceptable on a desktop computer. Some users would appreciate a 2.1 port to connect a display without taking up the thunderbolt ports, and it is borderline unacceptable that a new machine with a powerful chip in 2022 does not come with the latest HDMI port spec. Apple should be ashamed. /endrant. Kinda. I wish Apple made the giant Apple logo on the top of Mac Studio a MagSafe charging spot. It’d be perfect to be able to charge your phone on top of the Mac Studio, but I digress. Look, for most people, Mac Studio’s port selection will be more than enough. Factoring in a display like the Studio Display or the UltraFine, the I/O will be abundant and plentiful. I realize both of those are synonyms. Shut up. But the fact of the matter is that this computer is NOT made for most people; the people who are seriously considering Mac Studio need a larger selection of more, high-bandwidth ports — and that includes USB-A. Apple should have reconsidered the ports selection here because frankly, it’s not enough. Why am I ranting on this so much? Well, it comes back to the point I made earlier — I’m writing this review as a creator for creators. And as someone who creates things, this is a very annoying nitpick that no other review has mentioned. Pros shouldn’t have to tolerate sub-par ports on a $2000+ computer. Nobody should. Apple could and should have done better.
Another thing that I’ve seen talked about ad nauseam is Mac Studio’s thermal envelope — specifically its sound profile. I’d like to talk about this because there’s a lot of misleading information out there. I put Mac Studio’s thermal system to its limits — I ran Cinebench R23 for 30 minutes, Wild Life XTRM in Unlimited mode, and Compressor batches non-stop. Here’s what I learned; Mac Studio is kinda loud at idle, with a caveat. If you’re in a perfectly sound dampened room with absolutely no ambient noise, you absolutely will hear the fans. Instead of a high-pitched whining, it’s more of a dull sound. But alas, you can hear it. Though, this isn’t a big deal, especially if you have the Mac Studio a couple of inches away. What is a bit strange is that I can feel air coming out of the exhaust vents at the back of the machine pretty much all the time — including when the machine is in sleep mode. Odd. But it gets weirder than that — the air coming out is pretty much always slightly cooler than ambient. I’ve done a bunch of research over the past 2 weeks, and I haven’t been able to find out why other than the fact that a fan creates the effect that the space is cooler. Sure, it’s not that much of a difference, but for the amount of performance this machine offers, it’s slightly off-putting to see thermal results like this. It’s good to keep in mind that I don’t have the M1 Ultra variant, and as a result, don’t have the copper cooling system — though teardowns show that the M1 Max does have a little copper boat that sits on top of the exposed RAM. Look, whatever the case is, I’ve seen this “noise” issue talked about way too much so here’s the TLDR; it’s not very loud in any other place except for noise-isolated environments, and it blows cold air out the back. And well, that’s pretty much the story of every Intel Mac ever made, so I digress. When pushing Mac Studio’s thermal system, I got surprising results. Loading up Cinebench R23 for 30 minutes in multi-core, iStat reported temperatures of 38 degrees Celsius at their highest, and 27 at their lowest — barely more than ambient sitting at 23.8. Granted, I wasn’t pushing the CPU and GPU simultaneously, but from what I’ve been able to see, Mac Studio handles heat very, very well. And it should for having 2 massive blower-style fans and a giant heat sink at the back. From just putting my hand on the machine during testing, the enclosure stayed cool — cooler than my office. But, what changed was the amount of heat I felt at the back — it was JUST above ambient, which is pretty impressive considering I was firing all 10 of the cores at max load. The fans sat at a constant 1500–2000 RPM, which is pretty impressive since at idle, the fans sit at 1300. With all this information, you might be asking yourself if throttling was a bottleneck — and I can say confidently that it wasn’t. Anecdotal evidence proves that the machine never slowed down during the test — it was something I’d never seen before. The machine was never thermally bottlenecked and had ample room to go which leads me to believe that if I were to simultaneously push the CPU, GPU, and media encoders, I’d have miles of thermal headroom to go. It’s insane. One more thing before I wrap up this topic — Mac Studio with its larger thermal system destroyed my MacBook Pro with M1 Pro. Teardowns of the MacBook Pro show the same thermal system for M1 Max machines too — and I could feel the thermal envelope of the MacBook Pro struggling to keep up. If this was an M1 Max machine like the Mac Studio, I would expect for there to be severe thermal limitations. And not to spoil too much of what’s to come, but I can confirm that the MacBook Pro is thermally constrained — not by much, but more thermally constrained using a lower-TDP chip.
With Mac Studio being a powerful and ridiculously expensive machine starting at $2000 with no display, mouse, keyboard, microphones, speakers (okay, it does have a speaker but it’s hilarious), and camera, there’s no doubt that consumers are going to be keeping this machine for years. I know I will. Apple’s pro desktop market has suffered from infrequent updates during the Intel days, and I don’t see that changing with Mac Studio. There isn’t much of a need to do so. But this leaves one big problem — and that’s repairability and upgradability. As a machine ages, its fans get clogged up with dust, its power supply might fail, and its SSD might be borked — but how easy are these parts to fix and replace? iFixit gave Mac Studio a 6/10 on their repairability scale — and I don’t agree with them. For one, the PSU on Mac Studio is completely exposed. This is a huge hazard for anyone who decides to open Mac Studio up, and it is unacceptable that a machine in 2022 has an exposed PSU. Inside Mac Studio, only 2 parts seem semi-modular (except for the power supply, but good luck finding an OEM one) — the ports and the SSD. I’ll get to the SSD in a minute, but the ports seem pretty good. They’re not soldered like the iMacs, but rather come out with just a couple screws. Nice! But again, good luck finding components for Mac Studio — this is a very niche product that’s brand new, and it’s going to take months or possibly even years for 3rd party manufacturers to get parts. Here’s hoping Apple’s upcoming self-service repair program includes parts for Mac Studio. But now, let’s get into the drama bit — this SSD situation. YouTuber Luke Miani led the chase on this, and I’ll be using his research as the basis for my assumptions. Mac Studio “features” 2 slots for what seem to be proprietary SSDs. On models with 512 GB, 1 TB, and 2 TB SSD configurations, only one slot seems to be taken up. On models with 4 and 8 TB, it seems like both slots are taken up — the 4 TB model with 2 TB x2 and the 8 TB with 4 TB x2. This is most certainly because Apple can’t make 8 TB NAND chips at scale, so they’ve opted for using lower storage capacities twice and running them in RAID 0. Perfectly normal. However, upon further inspection, this is not the case. Well, the whole 2 slot thing is, but the SSD thing isn’t. An SSD needs 2 things to be an SSD — it needs a NAND flash storage chip on the PCB, and an SSD controller to talk to the NAND chip. On Mac Studio, only one slot on all configurations has an SSD controller, meaning that we couldn’t just swap in an SSD from another Mac with another capacity. The SSD(s) that come with the computer you ordered are tied to the SSD controller, which is soldered to the logic board. What does this mean? It means that Mac Studio’s storage is NOT upgradable, which is very very disappointing. What makes this worse, and perhaps infuriating, is that Apple could have included an SSD controller on the other slot of all Mac Studios, and in software, could remove the lock where the SSD that comes with Mac Studio is tied to the SOC. Apple has the power to do this, and their Geniuses can in-store. At the Apple Store, the Genius Bar has a “magical tool” to reprogram any Apple controller and assign it a serial number, meaning that if you got a logic board replacement, Apple could simply re-assign the new logic board your old serial number. If Apple made this tool publicly available and removed the ridiculous limitation which doesn’t allow the Genius Bar to make any changes to the serial number without a repair order OK’d by HQ, this wouldn’t be an issue. But alas, it’s not. So this leaves one more question — can you swap the same storage capacity from one machine to another, proving the Mac Studio to be repairable? The answer is… TBD. iFixit showed some luck in Apple Configurator when they swapped the NAND chips (remember that these are not SSDs), but I’m pretty sure the system SOS’d (which is hilarious by the way, check out Luke’s video on that) when it realized that the SSD didn’t match the SoC encryption key. sigh. So what does this mean for repairability? It’s a bit of a mixed bag. Logically, you should be able to swap NAND chips, but it doesn’t seem like the firmware wants that to happen. Here’s hoping that this is just an early-adopter bug and that Apple issues a firmware update and allows for NAND chip swaps in their Apple repair program because that’s all we’ve got. Pretty disappointing.
Now, for the moment you’ve all been waiting for — the part that Apple’s event hyped up the most, and the part that matters the most for pros and consumers alike. Let’s talk performance. Because I have so many benchmarks and tons of information, I’ll be splitting the performance section of this review into 7 parts; CPU synthetic benchmarks, GPU synthetic benchmarks, audio production, 3D modeling, photography, Compressor, and Final Cut Pro. This way, you can find the benchmarks applicable to your field easily. For example, if you’re into music production, CPU synthetic benchmarks and audio production might be helpful to you. If you use blender and other 3D modeling programs, CPU and GPU synthetic benchmarks and 3D modeling benchmarks might be important. And if you’re a developer, tough luck because I didn’t have enough time to test compile times in Xcode. They’re fast though — can confirm. Each section (and benchmark) will be accompanied by charts showing the benchmark name, what it’s testing, different computers, a scale, and what to look for.
Okay, enough intro business — let’s get talking; starting with CPU synthetic benchmarks. When I say CPU benchmark, you might think of Geekbench and Cinebench. You’d be right. Starting with trusty Geekbench, we see comparable CPU multi-core scores. It’s important to mention that because Geekbench doesn’t put sustained load on the machine, we’re not testing for thermals here. Looking at the data, the M1 Ultra Mac Studio takes the lead, but not as much of a lead as we were expecting. It only averages about 900 points higher than the 28-core Mac Pro. However, it also nearly doubles the score of the M1 Max Mac Studio. Here’s another interesting note — the M1 Max Mac Studio consistently performs higher than the MacBook Pro in this test. I can only attribute it to one thing and one thing only — High Power Mode on the MacBook Pro. The M1 Max MacBook Pro’s data was pulled from a 14”, with no High Power Mode. However, I assume that because there are no cooling and power draw limitations on the Mac Studio, the SoC is clocked higher, emulating High Power Mode. This could also be why we hear the fans constantly. The 12 core Mac Pro trails behind the $2000 base model Mac Studio, which is still hilarious to me, and right after that comes the base model 14” MacBook Pro. Honestly, pretty impressive results! I didn’t include the 8 core Mac Pro or any Intel MacBook Pros in this test because it would kind of be hilarious how the Mac Studio smokes those 2 machines. Goodbye, Intel! You’ve been so bad to us. One more thing — the Mac Studio took around 5 minutes to complete the test, while the M1 Pro took 6 and the M1 took 7.5. Interesting. Moving along to Cinebench R23, we finally start to see thermals kick in. This is one of the most important tests because it’ll show us how Mac Studio’s thermal system compares to the MacBook Pro. Anyways, we’re seeing pretty large gaps in performance between M1 Pro and M1 Max. Now, it’s important to know that this is the 8 core MacBook Pro — but adjusted for 20% scaling (2464.8 extra points), we get a score of 12,019.8 in the multi-threaded 10 minute sustained test — quite a ways away from the Mac Studio’s 12324. We can’t directly attribute this to thermals since it is an estimate in an otherwise perfect scenario, but based on the information we got from the Geekbench test, it’s safe to assume that there is some overclocking going on with the Mac Studio. Notice how I’m talking about the clock speed more than thermals — it’s because these 2 go hand-in-hand. Apple silicon is good enough that the chips don’t thermal throttle except in very limited scenarios. It’s more likely that the chips are clocked higher rather than the chips being thermally constrained in the MacBook Pro. Regardless, it’s true that the M1 Max in the Mac Studio performs better — even marginally — than the MacBook Pro. The Mac Studio even gets close to the Intel Core i9–12900K — a massive chip that could never perform in the chassis and thermal envelope of Mac Studio.
The CPU benchmarks so far support my hypothesis that the Mac Studio performs better than the MacBook Pro, but let’s take a look at GPU benchmarks. Once again, we’ll start with Geekbench Compute. Like Geekbench CPU, Geekbench Compute doesn’t push the SoC, so we can’t check for thermals or any other unique advantages. Unfortunately, Apple still hasn’t nailed down the GPUs on their Apple Silicon chips. M1 continues to struggle comparatively — the highest end M1 Ultra and Highest end M1 Max still can’t compete with comparatively specked Mac Pros. Yes, you could make the argument that Mac Studio is cheaper and smaller, but it’s still not able to compete with Mac Pro in raw synthetic benchmarks. Disappointing. But what’s not disappointing is that the M1 Max Mac Studio (both models) consistently performs better than the MacBook Pros. It’s not by the largest amount, but in metal, both Mac Studios outperform the MacBook Pros by about 4000 points. That’s also nothing to laugh at either — these are actual performance differences with real-world implications. If we look a little closer, the 24-core M1 Max Mac Studio comes eerily close to the 32-core M1 Max MacBook Pro in computing scores — thus proving that if you want raw power, the Mac Studio might be the more cost-effective option. One more little caveat — look a little closer at the 24 core M1 Max MacBook Pro and 24 core Mac Studio scores. Notice anything strange? Yeah, notice how the OpenCL scores for the MacBook Pro beat the OpenCL scores on the Mac Studio? This contributes to my second hypothesis about Mac Studio; Mac Studio doesn’t seem optimized for all graphics tasks right now. I’ll elaborate more on this in the Compressor section, but Mac Studio frequently lags behind the MacBook Pro in some tests. It’s really weird. Back to my hypothesis — it is kind of unfair that we’re using OpenCL to demonstrate this because OpenCL is deprecated in macOS, but we can’t ignore how the same chip in a MacBook Pro performs much better in an inferior body. Yikes. Another way that we can compare the thermal systems between the MacBook Pro and Mac Studio is Wild Life XTRM. This test stresses thermal performance. So, if we scale our results, we should see the Mac Studio smoke the MacBook Pro, right? Umm… no. You couldn’t be more wrong. When adjusted for 75% scaling, we get a score of around 14K — 4K more points than what the Mac Studio got. Eh? That can’t be right! No, it’s right. According to Max Tech’s MacBook Pro video last year, the 14” 24-core MacBook Pro got a score of 15K. Ummm… yikes! How could performance be… worse? Again, my hypothesis seems to be correct — there has to be something missing here, and I think it’s an optimization issue. This is a serious issue — why is the same chip performing significantly better in the MacBook Pro? This is very bad, and it’s slightly concerning to me that nobody else has reported on this. Is my Mac Studio faulty? Probably not, but it’s food for thought. Do you know what else is food for thought? Heaven! Not the heaven heaven, Heaven benchmark. This benchmark pushes the graphics cores and thermal system, and my MacBook Pro chugged. Thermal stability sucked on the MacBook Pro. The machine was almost too hot to touch, whereas the MacBook Pro remained cool as a cucumber. heh heh Ummm anyway, to my surprise, the MacBook Pro and Mac Studio performed similarly. sigh. Once again, the Mac Studio is not optimized for some reason. This time, I have fewer doubts about my machine being faulty — Max Tech also reported 94 FPS — the same as I got. If we adjusted the MacBook Pro for scaling… I’m not even going to get into that. The main theme of the graphics section is that something’s wrong with the Mac Studio. Whether that be optimization, firmware, or something else. I frankly don’t know. But something’s up, and something has to be done. I’ve been doing some anecdotal research and I found that others have seen similar results in benchmarks. My proposed solution is some sort of rewrite to the software for further optimization for graphics-related tasks — and OpenCL optimization. I’d also like to see updates to pro apps like Final Cut for optimization on M1 Ultra and M1 Max systems. TL;DR: graphics performance is good in a vacuum, but wails compared to the competition. I’ll touch more on this in the Compressor section, where I saw some interesting results.
Something that wasn’t as underwhelming as the graphics tests was the Logic Pro benchmark. This is the first real-world test and is pretty important if you’re into music production. The project (I don’t know what the terms are in Logic, so I’ll just use Final Cut terms like an idiot) is fairly simple; just some synthesizers playing in a loop. What we’re testing for here is how many tracks the system can handle before throwing up a system overload screen. Unlike Final Cut (well, if you have it turned off as any real pro would hehe), Logic tells you when the system can’t play the tracks anymore. I simply added tracks to push the CPU until the message showed up. This isn’t super important for people like podcasters since I’m working with synthetic music tracks, but this will be handy for musicians. Scaling was fairly linear, and I didn’t seem to hit any bottlenecks. The Mac Studio was able to play 122 tracks at the same time, the MacBook Pro was able to hit 118 tracks, and the iMac (which is just hilarious) was able to play 64. The 2 core differences between the M1 Pro and M1 Max accounted for 4 tracks — which is pretty insane. Apple doesn’t make a desktop with M1 Pro, so I wouldn’t be able to test for thermals — but this isn’t a sustained processor-intensive task. The SoCs barely stayed above ambient (aside from the iMac, which felt like it was going to light my house on fire). Playback was smooth, and the tracks continued looping for an entire night. Impressive. CPU usage was not high at all — it reminded me a little bit of a Geekbench test. It only stressed the CPU at the beginning of playback. After 10–20 seconds, everything worked like normal. It was almost like the computer wasn’t doing anything at all. This went for all machines, so I don’t think this is a real Apple Silicon advantage. If you are doing serious audio production at all, I wouldn’t recommend you go anything above an M1 Max. Moore’s Law seems to kick in by that point. I’ll get more into recommendations later but take note.
Unlike audio production, 3D modeling takes advantage of the cooling system a lot more. We already saw a lot of this in Cinebench R23, and how the Mac Studio (when accounting for scaling) took the lead. In the Blender benchmark, we see similar results. Let’s first start with CPU rendering, which is what many will be using (although not the fastest, it is the most preferred amongst artists). In the Classroom render, Mac Studio averages 10 samples per minute higher than the 8C MacBook pro When accounting for 20% scaling which we would see with the higher core count M1 Pro or M1 Max chips, we would get an average of 36 SPM. That’s 4 samples lower than the Mac Studio. Impressive — we’re not seeing those artificial bottlenecks come into play with the CPU as I expected. The same advantage applies to both the Junkshop and Monster benchmarks. When adjusted for scaling, Mac Studio performs slightly better in every test due to better thermals — awesome for Mac Studio buyers. On to the GPU test; this one was… actually pretty good. The Mac Studio beats both the MacBook and the iMac because of its vastly superior GPU. But, if we adjust for 75% scaling to emulate the scores of the M1 Max MacBook Pro, we get a score of 140 — just 2 points behind the Mac Studio with M1 Max. You might just leave it up to chance and the silicon lottery (and frankly, I don’t blame you), but if you look at these scores alongside all the other scores showing an improvement over the M1 Max MacBook Pro, the Mac Studio has a leg up because of (what I assume) it’s higher clock speed and better thermal system. These aren’t numbers to scoff at — these are real-world scenarios showing real-world benefits. You might be asking yourself “hey, why don’t we see any bottlenecks here like in the Wild Life XTRM synthetic benchmark?” Honestly, I wish I had an answer for you. These tests were done on the same day with the same computer — it’s really weird to see where the system lags. If there’s anything to take away from this, it’s that for 3D artists and people rendering in Blender a lot, the Mac Studio is an awesome machine. It’s fully optimized — and even faster than the MacBook Pro according to… math. Maybe the Mac Studio’s performance isn’t so bad after all…
Mac Studio has been designed for pretty much all pros — but I feel like photographers are going to particularly enjoy its benefits. You can pick any display, it’s powerful, and it’s small — allowing you to lug it around anywhere (if you can, thanks to the worldwide pandemic). As someone who spends a ton of time taking and editing photos, Mac Studio is simply awesome. I don’t have many numbers to back this up, since everything works in a matter of seconds (hence why this section is so short), but while doing the photography for this review, no matter how many layers I put on the image in Affinity (because Photoshop is dumb), everything flew. On previous computers, messing with the HSL tabs caused noticeable lag and beach-balling. That didn’t occur with the Mac Studio. However, it’s important to note that this isn’t a particularly demanding load. It’s just putting a couple of cores to the test. But regardless, there’s an actual tangible difference when working on complex projects. You’ll feel this sense of snappiness when moving things around. I know, this isn’t something particularly new with M1 Max — and to be honest, you probably don’t even need the M1 Max. My M1 Pro provides comparable performance and that same snappiness. However, exports just were faster, adding LUTs was instant… you get the idea. If you’re a photographer and strictly a photographer, you really can’t justify the Mac Studio. But, if photography is part of what you do, the Mac Studio is so fun to edit with. ML tasks, layers, LUTs, HSL, and all. It’s simply superb.
Now, onto my favorite part of these reviews; video production. Let’s start with Compressor because it’s usually what I and a lot of other video professionals use to start a project. If you don’t know what Compressor is, it’s an application made by Apple to encode video. For example, if my camera spat out H.264 footage and I want to edit in ProRes, Compressor will let me do that. Then, when I’m done and ready for final production, I can encode my ProRes video into H.264 for final delivery. It’s a GPU-intensive application that utilizes the media encoders and can be time-sensitive. There are a couple of things to note with Compressor (and any other video program); for one, Compressor results do NOT scale linearly. If the Mac Studio took 5 minutes to encode, we can’t expect the M1 Pro to take 10 minutes because there are double the media encoders/cores. This is not how video works. Two, not all video clips are created equally. A black screen that’s in 4K will encode much quicker than a scene with moving parts and changing ISO. That’s how video is translated into bits and bytes. It’s more complex. Lastly, each codec takes a different amount of time — H.264, because it’s heavily compressed, takes longer for the computer to encode than ProRes. Something weird with video is that the more compressed a video is, the harder it is to deal with. I lied; there’s one last thing to keep in mind — all tests were done with the new optimized version of Compressor that came out on April 12th. I completely scrapped my previous Compressor section. The text you’re reading right now is written at 1:45 AM on Wednesday, April 13. My previous results were garbage and irrelevant. All the subsequent graphs that you’ll see are also updated with new information. With all the prerequisites out of the way, let’s get into the results. I did 2 fairly common tests with Compressor — HEVC to ProRes, and ProRes to H.264. ProRes is a fairly uncompressed format, and you’ll see that in our results. HEVC is super efficient and also pretty uncompressed while retaining small file sizes, and iPhones record in this format. In the HEVC to ProRes test, the Mac Studio handily beat the MacBook Pro by about a minute — which is to be expected. Remember what I said earlier about results not being scalable? That strongly applies here. If you go out and find a Compressor benchmark for an M1 Max MacBook Pro, you’ll see that they don’t differ all that much from an M1 Pro or M1 — especially in a test where the media encoders do much of the heavy lifting. During this test, GPU utilization was only at about 20–30% because we were encoding to ProRes, and using 2 easy-to-work-with standards. Another important thing is that Compressor doesn’t make the computer heat up very much. Temperatures were barely above idle on both machines, and as consequence, the thermal envelope of both machines didn’t affect the results. What can we take away from this? If your job solely depends on encoding/decoding HEVC and ProRes footage (which is a specific niche), Mac Studio won’t give you earth-shattering results. You’re better sticking with an M1 Pro. The extra speed doesn’t matter. However, the same couldn’t be more wrong in the next test, where I encoded a ProRes clip to H.264 — a very lossy and compressed video format. This annihilated the GPUs while not being all too encode engine intensive. You might also think that by consequence, the thermal system would be a large determining factor here — but you’d be wrong. Both computers were super cool to the touch. With all that being said, the M1 Max Mac Studio whooped the MacBook Pro so hard that it’s actually kind of funny. But remember — we can’t account for 75% scaling here just because every scenario is different, so this is a pretty good result. The Mac Studio is fully optimized. Just don’t expect any massive performance gains comparing it to the M1 Max MacBook Pro. To wrap this whole Compressor section up, you might be slightly disappointed in the Mac Studio. But don’t be — these scores in a vacuum are EXTREMELY impressive for a computer. They make a laughingstock out of the Mac Pro — even MKBHD’s Mac Pro. The Mac Studio performs awesomely in this test after its optimization update — there’s practically no difference between it and the MacBook Pro. However, if you bought or plan on buying this computer specifically for this, I’d hold off and maybe consider an M1 Pro MacBook Pro. More on the recommendations in a bit, though.
Finally, it’s time for the moment all of you have been waiting for, the grand finale of the performance section, the crème de la crème of Apple silicon, it’s time for the Final Cut Pro test. Look, I’m sure many of the Premiere users might be disappointed since I’m not doing a PugetBench benchmark for Premiere, but suck it man. Final Cut Pro is better. Much like the Compressor test, there are a couple of things to keep in mind that I will shamelessly copy and paste: for one, Final Cut Pro results do NOT scale linearly. If the Mac Studio took 5 minutes to export, we can’t expect the M1 Pro to take 10 minutes because there are double the media encoders/cores. This is not how video works. Two, not all video clips are created equally. A black screen that’s in 4K will encode much quicker than a scene with moving parts and changing ISO. That’s how video is translated into bits and bytes. It’s more complex. Lastly, each codec takes a different amount of time — H.264, because it’s heavily compressed, takes longer for the computer to encode than ProRes. Something weird with video is that the more compressed a video is, the harder it is to deal with. I lied; there’s one last thing to keep in mind — all tests were done with the new optimized version of Final Cut that came out on April 12th, as well as the latest Final Cut Pro supplemental update package and Pro Video Formats. With all that said and done, let’s get into the benchmarks. First, starting with the render test. This project is fairly simple — 4K iPhone footage in H.264 straight from Compressor with the “Boogie Lights” text applied and a custom, yet light LUT applied. Rendering took around 3 minutes longer on the MacBook Pro than on the Mac Studio. Now, before the pathetic Apple hater-losers get their boxer briefs up in a twist with their bogus nonsense (you know who you are), remember what I said earlier? Video projects do not scale linearly — especially with encode/decode blocks. Benchmarks from Luke Miani’s 16" MacBook Pros show that the M1 Max MacBook Pro (32C) only performed 2–3 seconds faster in his render test than the M1 Pro (16C). I know this isn’t the same project, but if we did this test on a MacBook Pro, we’d only see about a minute or a minute-and-a-half gain. The rest is all due to thermals. Because rendering pushes GPUs to their extremities, both computers heat a ton. I would be willing to bet that at least a minute here is due to the thermal differences between the MacBook Pro and Mac Studio. I can’t say this for a fact, but it looks like it. Whatever the case may be, the Mac Studio is well-optimized in Final Cut, and differences will be apparent between the Mac Studio and MacBook Pro. Breathe easy, Mac Studio video editors. However, this test is already pretty irrelevant — Final Cut Pro renders in the background. What’s more important is how the timeline feels when you’re editing. When I was on Intel, I never dared to switch from “better performance” to “better quality.” This is the first time I’ve done that, and while scrolling through this timeline, things just fly. It’s truly outstanding. Yes, you could get this with M1, but after the new update, it’s just buttery smooth. I haven’t encountered a single beachball — it’s great. But the (second) most important performance metric in Final Cut is export times. Since M1, gone were the days of waiting all night for your computer to export an extremely important video just for it to fail. I’ve cried because of days like these. Export times on the Mac Studio are super impressive, but before I dig in, here’s how to interpret this data — the computer heats up substantially during export. This means that the thermal system plays an extremely large role in export times. It’s why all the YouTubers made all that noise about the 2018 MacBook Pro’s i9 thermal throttling issues. The export test also relies heavily on the media encode/decode blocks on Apple silicon — loosening the reliance on the GPU. GPU usage on the M1 Pro was about 44% instead of 100%, and on M1 Max was 27%. The encode/decode blocks matter more here. However, this does not mean we’ll see 2x performance improvements between computers (or more, because thermals) because the GPU eats up the leftovers that the encode/decode blocks leave behind. With all that said, the Mac Studio whooped the MacBook Pro in both thermals and performance. The MacBook Pro went up to 62 degrees celsius, and the Mac Studio stayed at 48 degrees. So, this should mean that we should see massive performance enhancements, right? I’m happy to report that you’d be correct. The Mac Studio exported the same project 33 minutes faster than the MacBook Pro, a huge performance improvement. It’s hard to tell how much of this is due to thermal enhancements, but it’s a huge improvement no matter how you phrase it. If you’re a video editor, Mac Studio is going to be such an amazing purchase. I’m glad to see that Apple fixed its huge performance issues in their pro apps. I can confidently recommend Mac Studio for all video editors — and that’s huge.
To wrap up this performance section, I can only say a couple of things — if your work requires good CPU performance, Mac Studio is simply the way to go. It’s more powerful than the M1 Max MacBook Pro in tons of ways, it’s customizable and has tons of perks. However, if you’re using anything other than Apple’s Pro Apps like Logic, Compressor, and Final Cut to perform graphics-related tasks, I can’t recommend Mac Studio to you just yet. Something is going on with the software optimization that makes it fall behind compared to Apple’s other Macs. Here’s hoping for software enhancements in the coming weeks for macOS to take full advantage of Mac Studio’s performance. And that about wraps it up for the huge performance section of this review.
With the massive performance section out of the way, you probably have tons of questions as to which computer you should get. Don’t worry, I’m here to help and interpret this data. I’m going to break it down into workflows: music production, video editing (light), video editing (heavy), video editing (extreme), photography, 3D modeling, software development, and others. If you create music on the daily or edit podcasts, I recommend the 24-core M1 Max, with 32/64 GB of RAM. You don’t need 20 cores, and you don’t need the extra GPU power. As for RAM, it depends on how many tracks you play with. If you do extremely heavy music production, go for the 64. If not, unified memory is super efficient, and you’ll probably be fine. What about if you’re into 3D modeling? I’d recommend the same config. Though some may find the extra GPU performance handy, you probably won’t see a difference in the real world. You’ll also probably be fine with 32 GB of RAM, but go big on storage. How about software development? I’d recommend the same spec, but with as much RAM and storage as you can afford. You’ll need it. Compile times are great on these machines. Onto video editing. For light video editing, I wouldn’t recommend a Mac Studio at all. You’re much better with an M1 Mac mini. If you do want a Pro product, I’d recommend a refurb MacBook Pro 14”, though you probably don’t need it. Go big on storage, not on RAM. And get the M1 GPU upgrade. If you do moderate/heavy video editing, it’s up to you if you want to go for the 24 or 32 core M1 Max. Performance deltas will probably be very minimal, but if you plan on keeping the computer for a while, I’d recommend the 32 core if you see your needs changing. Also, spend your money on storage, not RAM. I’ve never run into a RAM issue while editing. Finally, if you’re into heavy video editing, I’d recommend the M1 Ultra just for that extra oomph. You won’t see massive, ballistic differences like some of the Apple fanboys say, but it’s worth it. I’d personally go 64 GB of RAM and 48 core GPU, but if you are going to spend more money, spend it on storage and GPU upgrades. If you’re doing anything else like artwork or graphics, I’d recommend the base model. It’s honestly the best spec for the value. It’s an incredible machine with an incredible value proposition.
Mac Studio is a different kind of computer — it’s not just something the average person can just pick up to have a desktop Mac in their house. It’s a tool designed for a very specific subset of people. Coming back to what I said earlier in the intro, Mac Studio was made for the enthusiast who wanted a Mac that could just get stuff done. They didn’t care much about the customizability or whatever — they wanted a sleek and small yet portable computer that just ripped through everything. So this is for you, Mac enthusiasts. Apple did it. They’ve made a truly amazing tool. And that’s something no other computer company can ever do.
This review was produced following 3 weeks with the Mac Studio (2022) base model purchased by me on launch day. Of course, Apple had no editorial impact on this review, and they’re seeing it for the first time you are. Stay safe out there and mask up when around others, and let’s get through this together. Stay home, stay safe, and stay healthy — and I’ll see you around.
