Were you SHOCKED by the SECRET we found in the M1 Max Mac Studio cooling system? Comment below! NEW M1 Ultra Chip T-Shirt ➡ max-tech-store.creator-spring.com/listing/apple-m1-ultra-soc-package Like our Wallpapers? Download them here! ➡ bit.ly/2WNc6Qw Download TG Fan Pro for Mac ➡ geni.us/dkbrc75 iFixit toolkit ➡ geni.us/p8JVj0 Best deals on M1 Macs on Amazon ⬇️ Apple Mac Studio with M1 Ultra ➡ geni.us/cgkl Apple 27" Studio Display ➡ geni.us/GeA3te6
I been watching your videos for months! Each video is HIGH Quality and from the thumbnail I could easily see that there was a LOT of difference between the aluminium and copper heatsinks/cooling. Thanks for covering the reason behind the difference in weight. Love your videos, always! Watching this at 2:12AM lol!
Copper has almost double the thermal conductivity for the same area. The reason the temps are the same is that the M1 Ultra occupies a larger (almost double) area. So it all evens out. Idk why I’m pointing this out I guess it’s just nice to see that the results are consistent with physics.
Thermal conductivity is an intrinsic property of the material and not related to area. For a heat sink, aluminum alloy is usually used because you can increase the cross section to achieve the same thermal resistance in plane as copper. If you have to use copper it’s because there was no space. Copper actually has more than twice the thermal conductivity of the aluminum alloys used here, since pure aluminum is rarely used.
@@justinhalsall4077 the radiative coupling will be relatively weak if the absolute temperatures and their differences are relatively low. Also note that in the infrared wavelengths of interest the colour is not important, it’s the thickness of the paint that dominates the emissivity, since if it’s too thin the paint layer becomes transparent and the metal underneath has low emissivity due to the electron mobility
@@ProfessorHamer maybe I have the scientific terminologies in English wrong but I do know that the thermal properties and area covered and volume used are all correlated. Thank you for adding to my knowledge.
at 11:01, you can see in the top right (nearish the clock) the gpu is running at 100% while the cpu is at about 24%. This suggests to me that running the two tests at the same time meant the cpu wasn't able to feed the gpu fast enough to max it out.
Hey Max, congrats on your teardown. You have everybody in the industry using your images and talking about you, hopefully you'll get a few more hundred subs. I know business is hard and you always gotta be ahead of the pack, well this time you tore it up lol. Please add Illustrator or Coreldraw to your benchmark test since a lot of us graphic artist are also deciding which Mac to go with. Great Job as always. Blessings 🙏🏾
Also, photo editing programmes such as Photoshop, Affinity Photo, or Luminar Neo. It would be great to see these results. Can you supped up the graphics core on the studio? Havagooday Greg
@@jabulaniharvey yup. Sometimes I have 100mb files in coreldraw and illustrator and I'd love to see how the M1s respond to those kind of files especially when you still have much work to do with them.
The ULTRA has a vapor chamber in place of the heat pipe in the MAX (at 4:30). The vapor chamber is similar to a heat pipe in terms of principle, but it is constructed by layering and fusing copper sheets instead of starting from extruded copper tubes. The advantage of the vapor chamber is that the enclosed volume isn't limited to one dimension (the tube profile), so heat can spread in two planes instead of one.
The M1 Ultra heatsink fan (HSF) had a copper vapour chamber to transfer heat from the M1 Ultra heatspreader to the copper finstack. The M1 Max HSF features a more simple and conventional copper heat pipe to aluminium finstack setup.
The different thermal pastes (thermal interface materials) is likely due to several factors: There is a requirement to minimize resistance to heat flow (thermal impedance) and also a need to accommodate tolerances of a variety of chip components within the multi-chip module. Since the CPU is the most power dense, the engineers want to minimize thermal impedance path from the CPU chip to the heat sink system, thus the thermal interface for the CPU is optimized for the lowest thermal impedance (the gray stuff). Since it would be more costly to make the memory chips perfectly level with the CPU (eg. precision grinding/lapping of an already-assembled SOC module) and use a single plate (as they did for the ULTRA), the engineers opted instead to use another type of thermal interface material for the memory chips (the black stuff). This is actually a gap filler material since the mechanical limits of the CPU interface will actually limit how close the memory can get to the heat sink. Gap fillers are designed to be highly compressible with low application force and then stay in place. The gap filler for the memory probably has a much higher thermal impedance than the CPU because it is thicker, but that is OK because the memory also has much lower power density than the CPU.
Nice Video! Quick comment one your statement from 12:45 ("Apple is limiting graphics performance"). It's very likely that they are not. The Geekbench stresstest is very demanding, taking up almost every tiny bit of the CPU, leaving not much for other tasks (like moving the mouse or opening a program etc.). Even though the 3D mark stresstest tests the GPU, it also needs some resources of the CPU. So because the CPU is already so busy with Cinebench, it probably can't feed the GPU fast enough for the 3D mark test. Long story short: With Cinebench running, the 3DMark stresstest is CPU limited. At least that is my humble opinion on the matter. Maybe worth trying another combination of CPU and GPU stresstest.
Yeah the second he made that statement I knew he made a large mistake with his logic, Apple is not putting any artificial limitations for this scenario to happen, it simply becomes the CPU becomes the bottle neck with the specific tests they are running.
I was going to comment the same. GPU test can’t run at 100% if it was using 25% CPU alone and Cinebench is stealing most of those cycles in the combined test.
@@MoonshineOctopus We have been seeing bottleneck tests for decades now on systems, I don't think the reviewer in this case is as familiar with this, it is difficult to get a test to stress 100% of both CPU and GPU as it is perfect balance needs to be reached and that balance differs on every system config.
Seriously great content on a consistent basis - you guys are killing it. Articles across the net referring to your work taking the Studio apart! Deep dives and analysis on another level. Max Tech is the gold standard..
Please run a simultaneous session of: 1. Xcode compile test (repeatedly over and over); 2. Logic w/ multiple tracks of white/pink/brown (different types of randomised) noise + plugins; 3. Video editor with a 4k edit. Time the compile tests. Track that time against the # of tracks in Logic as you increase them and the size and complexity of the edit. Artificial benchmarks are fine, but there is no real substitute for real-world workloads.
*Is there a high power mode in the Mac Studio like there is for the MacBook Pro? Maybe if you select it, it will enable both the CPU and GPU to run 100% at the same time. Try this on the MacBook pro aswell.*
Love your energy for these videos. Still waiting for mine to arrive. I got the M1 Max with 10-core CPU, 32-core GPU, 16-core Neural Engine 32GB unified memory 1TB SSD storage. Figured this should be more then enough power for me for hopefully years to come
Years to come? You wont be needing another computer DECADS to come unless apple decides to start slowing down their chips like they did in the old days lamo
@@samuelchimezie I still use a MBP 2012 fully upgraded and works great at a decade old. So this will will be a huge upgrade. I am thinking of getting the Studio display as well. Then see what comes out in the next year or so and get another MBP with M2 or whatever chip is avail then
the coverage on these products on your channel compared to even some of the biggest reviewers in the GAME is groundbreaking and it should be recognized fr
Honestly, I really had thought that the Studio was a bloated Mini with a big chip in it. Your tear-downs have completely eradicated that thought. I don't know if Apple appreciate you or not (probably not as you specialise in leaks!) but they should be appreciative of these Studio videos because they reveal the massive amount of effort and thought that Apple has put into creating these models. These units are so technologically dense that I think they are worth what is being charged and, by virtue of your videos, we now know that Apple could very significantly increase the performance of these machines any time they wish with a simple software update that allows another 20degC or so for the chip. Great job, guys!
Really impressed that you go to the Tear down path to show their differences. Keep it up and this will surely help potential buyers with their decision^ Hope for that 1M subs soon
The supposed full copper heatsink of Studio with the M1 Ultra is most likely using a vapour chamber from what you showed. And not surprising given the size of the M1 Ultra as compared to the M1 Max. But great in-depth look at the coolers and cooling capabilities.
7:01 the M1 ultra have a much larger heat pipe than that M1 Max. the heat pipe is not made of pure copper, infect they pulled almost all the air out of the pipe and added some water vapour in it, making it extremely efficient in transferring heat than copper by condensation. it is worth noting that this structure does not necessarily need to be done in a pipe looking way, in fact, it can be done in all different shapes as long as the condensed vapour can return to the chip efficiently so that it can be heated up again ( to transfer heat ), it will work. I expect the entire "Copper" instruction of that area covering the chip to have empty space ( low pressure when compared to 1 atm ) and contains some water vapour such that it will function like a heat pipe. since heat pipe is much more conductive than pure copper can is capable to spread heat much faster, I can't think of any reason apple will not that the design described above.
Hm, maybe the gpu limit is not because of power and temp, but instead because cpu can’t control the GPU for the test at full speed while also being under full load on its own
It doesn't necessarily mean that it is Apple who limits the gpu performance when the cpu is maxed out. Its quite possible that something has to be prepared on cpu to provide the workload for the gpu. If cpu is maxed out, it becomes the bottleneck, and it doesn't provide enough stuff to max out the gpus.
I notice that in second SSD port at M1 MAX there are controllers that are missing in second port at M1 ULTRA which course on tests that second port could not be use for upgrade. Because there are chips in M1 MAX that would suggest possibility for upgrade. Can you please check that ?
UPDATE to my original comment!! I Isolated out an issue that was causing the slow export times on both the Mac Studio and the Intel MacBook Pro. It turns out that if you add the "Outline" effect to a title the export becomes more than twice as long for some reason. So the Mac Studio with M1 Max is now twice as fast as the Intel MacBook Pro I was comparing it to. The M1 Max now exports the same 6:30 minute clip in 53 seconds and on the Intel MacBook Pro it now takes 1:10 minutes instead of 4:23 minutes. Super happy I found this out! My recommendation to anyone using Final Cut Pro is not to use the "outline" effect on a title if at all possible otherwise your export times will be much longer. So far I am disappointed with H.264 performance on my Mac Studio with the M1 Max. In nearly every test that I've done the export times in Final Cut Pro are always about 36 seconds slower than my 2019 i9 15 inch Intel MacBook Pro. My expectations were super high for this but seeing as it is slower than older Intel machines when it comes to H.264 footage I may end up returning it until a fix comes to change this.
So your $2000 machine is 36-seconds slower than your $3000 machine while being completely silent and the fans don’t kick on if you connect two displays? Ehhh return away, I guess… someone else will snap up your folly of a refurb.
@@Wizzy0037 I actually have an update to this original comment, I forgot to edit it. I ended up Isolating out an issue that was causing the slow export times on both the Mac Studio and the Intel MacBook Pro. It turns out that if you add the "Outline" effect to a title the export becomes more than twice as long for some reason. So the Mac Studio with M1 Max is now twice as fast as the Intel MacBook Pro I was comparing it to which is what my initial expectation was. The M1 Max now exports the same 6:30 minute clip in 53 seconds and on the Intel MacBook Pro it now takes 1:10 minutes instead of 4:23 minutes. Super happy I found this out! My recommendation would be to never use the "outline" effect on a title unless it is needed otherwise the export times will be much longer.
Do you ever sleep? You have done so many videos on all aspects of these new computers, it's amazing! Thanks for all the effort as the videos answer a lot of questions.
Seeing these results, I wonder if the chips are still operating as if they were inside something other than a Mac Studio? Thankfully, this is certainly something that can be corrected with software updates as the hardware design looks solid and there’s ample power and cooling on hand to crank performance higher.
1:45 In Luke Miani's trying to upgrade video, you could see that the SECOND SSD port had a BLANK PCB under the SSD port.. but here that isn't the case? Wonder if YOU try to upgrade the same way, will it work...
I dunno, the $1,999 one with 32-cores seems like the best bang for buck but this video helped me understand (on a hardware level) why that price isn't as crazy a jump. Neat video dudes, you guys were the first to teardown the studio I think.
Got a question, do softwares and programs need to optimise for m1 ultra over the m1 max? Will apps need to update to use the extra power that comes with the chip?
I think that watt measurement is not a good indicator on the GPU usage. At a given frequency, watts indicate how many internal circuits are used. If a software does not use all functions of GPU because it doesn’t need , the watt will not be maxed but the usage will be at its maximum. The clock speed is to my point of view a good indicator of a potential throttle.
You can’t tell if the aluminum is doing as well as copper as it’s cooling a different chip. To really test the difference you’d have to toss the copper heatsink into the aluminum system and Vice versus. Then test the differences.
I only speak for myself, and I know you guys are getting re-blogged by the verge and all the other big sites... so amazing on that, but as a general user, I want speed tests to help with my purchase. I want performance tests on different types of footage. I want what it's like to be in a timeline. The tear down is a whole other level of user, that I think is only a small portion of the creative users looking to buy this machine. All of this is to say that I'm eagerly awaiting the performance tests of the max vs the ultra. And PLEASE, every TH-cam channels is testing only in FCP, let's see Premiere and After Effects stuff too please! Love the channel, keep up the great work. Been telling everyone to stay tuned to your channel for these numbers. Been with max since the camera review days.
Hey! i've been hearing rumours about a constant hum fan noise at idle state, on the M1MAX, but not on the ultra. This video doesn't seem to agree with it, but what do we know about it ? seems that people are complaining out there, I'm planning to buy the M1 MAX Mac Studio for Music Creation. I am very sensitive to noise, so a fan noise problem would kill me literally - thanks in advance if you read my comment !
Sounds like they are putting it someplace like a metal shelf, or a hollow plywood desk that would act as a drum head for the vibrations of the motor. Stick to real wood, and you won't regret it. Particle board and mdf is for cupboards, and not even good at that. As for metal, sit on an empty oil drum and fart. You'll hear it further than a lion's roar. Kids bang on metal pans for a reason. Acoustics, never forget acoustics when setting up your PC layout.
Thanks for this video Max, incredible work! Please could you comment on the noise of the Max vs the Ultra - people on forums (and also TH-camr ArtIsRight) are complaining that the Max is loud at idle (louder than the Ultra). Please could you put this issue to rest - are they the same?!
@@MaxTechOfficial Thank Max Tech, some people complain about the nature of the sound (on reddit and MacRumors), the Max seem to resonate more than the Ultra and add some unpleasant frequency at the same 1325 rpm speed. For audio mixing and recording, that can be a bummer!
A bit late commenting on that, but Apple isn't limiting the GPU performance in that last CPU+GPU stress test scenario. Graphics pipeline actually needs some CPU resourses for draw calls for the GPU to do its job, and if you max out the CPU resourses with a 100% stress test, it surely starts bottlenecking the GPU performance. Nothing to do with "wattage" limitations or any other limitations for that matter. Hope that clears it out.
I tried to break Final Cut Pro and I am sure if it ran on windows vista at the time it would crash or blue screen but it never crashed even on a hot i7 MacBook Pro from 08 . Apple is the gold standard of tech
I noticed you said it will be completely silent is that accurate? I’m an audio engineer and I often track vocals in the control room With my 2018MBP I hear fan noise I would consider getting a Mac Studio strictly off the fact that it is indeed silent if it actually is!!🙏🏻🙏🏻
GREAT Video... Thank you! I'm looking for an Adobe after effects ram preview and export comparison between the studio M1 MAX vs. the M1 ULTRA. I'm sure that video would get a nice chunk of traffic.
Excellent video. This is exactly the information I was waiting for to see if the Max Studio ran quiet when under stress like the M1 Mini does. Thank you :)
Nice informative video. Wonder if the GPU is being throttled when the CPU is at full power, due to limitation of target power draw of the M1 Max SoC. Similar to how the PS5's SoC architecture works with AMD SmartShift
Honestly I don't know which Mac Studio to buy. I sold my iMac 27" i9 from 2020 for this haha. I do video and photo editing (no 3D stuff) with sometimes quite a lot layers. What do you think?
Great stuff guys. When are you going to try swapping an SSD from one Studio to another? and also trying to add an extra SSD from one Studio to another?
btw I think you account has been spoofed as I just received a message from someone pretending to be you, asking me to message them on Telegram, and offering me an iphone and watch for free...
The power supply is 370W with the Mac studio. Both the Ultra and Max don't use nearly as much power. I wonder if Apple gave the Mac Studio that power supply for future models going forward?
This Apple choice of not ramping up the fans is weird for a pro machine that supposed to be designed to performance all the way. If you wanna keep at 60°C you ramp up the fan at the max. When the fan is at top speed, then you start to throttle the chip performance. I was wondering if there is any possibility that this machine isn’t on “Performance Mode” and that is why it’s behaving like that?
Great breakdown as usual. Just want to point out that in the combined test, you might be seeing some CPU limitation of the GPU because the CPU is already maxed out by the Cinebench app. I noticed that when you ran Wildlife Extreme by itself, there was about 24% CPU utilization, so it seems like that test does require some CPU to fully push the GPU to its limits.
Hey Max, Great highlight of the Thermal throttling with CPU + GPU peged, I'm wondering since you've TGPro, please do test running benchmarck with TGPro forcing the fans to the Max (yes that'll be loud but that's for science and people that don't care about noise)
I’m probably watching more than 80% of all the content about Apple Silicon on the internet, and let me tell you guys are on another level ! So glad to learn the Studio still has a dual fan with M1M, I wonder how it would compete vs a MacBook Pro with same specs…
So I'm confused. It's not thermal throttling but still throttling? I've been leaning toward the Studio for the issue of not thermal throttling but if still throttling seems I might as well go for the Macbook??? Purpose: music production but I could really use the portability.
Once again this video confirms that you are providing the BEST reviews of the new Mac Studio in the internet! This time, I am confident you will reach 1M fans very soon! Thanks for this review since this is the model I ordered because of your recommendation. Keep up the great job!
Were you SHOCKED by the SECRET we found in the M1 Max Mac Studio cooling system? Comment below!
NEW M1 Ultra Chip T-Shirt ➡ max-tech-store.creator-spring.com/listing/apple-m1-ultra-soc-package
Like our Wallpapers? Download them here! ➡ bit.ly/2WNc6Qw
Download TG Fan Pro for Mac ➡ geni.us/dkbrc75
iFixit toolkit ➡ geni.us/p8JVj0
Best deals on M1 Macs on Amazon ⬇️
Apple Mac Studio with M1 Ultra ➡ geni.us/cgkl
Apple 27" Studio Display ➡ geni.us/GeA3te6
I been watching your videos for months!
Each video is HIGH Quality and from the thumbnail I could easily see that there was a LOT of difference between the aluminium and copper heatsinks/cooling. Thanks for covering the reason behind the difference in weight. Love your videos, always! Watching this at 2:12AM lol!
You should install the M1 Ultra heatsink into the M1 Max motherboard, if its compatible of course!
Can it be overclocked? Does it need a firmware change or something to overclock?
What happened to the Fan Noise & Speeds section? What are the sound differences?
What about the fans of the MS M1 Max being audible when idle? There are some rumors about this on the interwebs. Can you confirm this?
Copper has almost double the thermal conductivity for the same area. The reason the temps are the same is that the M1 Ultra occupies a larger (almost double) area. So it all evens out. Idk why I’m pointing this out I guess it’s just nice to see that the results are consistent with physics.
Maybe with extra headroom if you load it up with max amount of RAM
Would the black paint help with the cooling too?
Thermal conductivity is an intrinsic property of the material and not related to area. For a heat sink, aluminum alloy is usually used because you can increase the cross section to achieve the same thermal resistance in plane as copper. If you have to use copper it’s because there was no space. Copper actually has more than twice the thermal conductivity of the aluminum alloys used here, since pure aluminum is rarely used.
@@justinhalsall4077 the radiative coupling will be relatively weak if the absolute temperatures and their differences are relatively low. Also note that in the infrared wavelengths of interest the colour is not important, it’s the thickness of the paint that dominates the emissivity, since if it’s too thin the paint layer becomes transparent and the metal underneath has low emissivity due to the electron mobility
@@ProfessorHamer maybe I have the scientific terminologies in English wrong but I do know that the thermal properties and area covered and volume used are all correlated. Thank you for adding to my knowledge.
at 11:01, you can see in the top right (nearish the clock) the gpu is running at 100% while the cpu is at about 24%. This suggests to me that running the two tests at the same time meant the cpu wasn't able to feed the gpu fast enough to max it out.
Hey Max, congrats on your teardown. You have everybody in the industry using your images and talking about you, hopefully you'll get a few more hundred subs. I know business is hard and you always gotta be ahead of the pack, well this time you tore it up lol.
Please add Illustrator or Coreldraw to your benchmark test since a lot of us graphic artist are also deciding which Mac to go with.
Great Job as always.
Blessings 🙏🏾
Also, photo editing programmes such as Photoshop, Affinity Photo, or Luminar Neo.
It would be great to see these results.
Can you supped up the graphics core on the studio?
Havagooday
Greg
At last! a commenter who requests tests for us illustrators!
@@jabulaniharvey yup.
Sometimes I have 100mb files in coreldraw and illustrator and I'd love to see how the M1s respond to those kind of files especially when you still have much work to do with them.
I wanna see a Mac run STEAM and the Quest VR software! Quest needs a lot more Developers! 😬
5:30 AM? That's Max dedication! Who am I kidding, that's ULTRA dedication! Thank you!
lmao
There is a thing called time zones…
@@adhprakash He's talking about 8:42 when Max was testing the CPU thermal stress at 5.30 in the morning.
The ULTRA has a vapor chamber in place of the heat pipe in the MAX (at 4:30). The vapor chamber is similar to a heat pipe in terms of principle, but it is constructed by layering and fusing copper sheets instead of starting from extruded copper tubes. The advantage of the vapor chamber is that the enclosed volume isn't limited to one dimension (the tube profile), so heat can spread in two planes instead of one.
The M1 Ultra heatsink fan (HSF) had a copper vapour chamber to transfer heat from the M1 Ultra heatspreader to the copper finstack. The M1 Max HSF features a more simple and conventional copper heat pipe to aluminium finstack setup.
The different thermal pastes (thermal interface materials) is likely due to several factors: There is a requirement to minimize resistance to heat flow (thermal impedance) and also a need to accommodate tolerances of a variety of chip components within the multi-chip module. Since the CPU is the most power dense, the engineers want to minimize thermal impedance path from the CPU chip to the heat sink system, thus the thermal interface for the CPU is optimized for the lowest thermal impedance (the gray stuff). Since it would be more costly to make the memory chips perfectly level with the CPU (eg. precision grinding/lapping of an already-assembled SOC module) and use a single plate (as they did for the ULTRA), the engineers opted instead to use another type of thermal interface material for the memory chips (the black stuff). This is actually a gap filler material since the mechanical limits of the CPU interface will actually limit how close the memory can get to the heat sink. Gap fillers are designed to be highly compressible with low application force and then stay in place. The gap filler for the memory probably has a much higher thermal impedance than the CPU because it is thicker, but that is OK because the memory also has much lower power density than the CPU.
Nice Video! Quick comment one your statement from 12:45 ("Apple is limiting graphics performance"). It's very likely that they are not. The Geekbench stresstest is very demanding, taking up almost every tiny bit of the CPU, leaving not much for other tasks (like moving the mouse or opening a program etc.). Even though the 3D mark stresstest tests the GPU, it also needs some resources of the CPU. So because the CPU is already so busy with Cinebench, it probably can't feed the GPU fast enough for the 3D mark test. Long story short: With Cinebench running, the 3DMark stresstest is CPU limited. At least that is my humble opinion on the matter. Maybe worth trying another combination of CPU and GPU stresstest.
Yeah the second he made that statement I knew he made a large mistake with his logic, Apple is not putting any artificial limitations for this scenario to happen, it simply becomes the CPU becomes the bottle neck with the specific tests they are running.
I was going to comment the same. GPU test can’t run at 100% if it was using 25% CPU alone and Cinebench is stealing most of those cycles in the combined test.
Agreed, try setting Cinebench to lower priority and see if that changes the results.
@@MoonshineOctopus We have been seeing bottleneck tests for decades now on systems, I don't think the reviewer in this case is as familiar with this, it is difficult to get a test to stress 100% of both CPU and GPU as it is perfect balance needs to be reached and that balance differs on every system config.
Came to say the same thing. This is almost certainly the case.
Since you have 2 studios have you consider taking ssd of 1 and put into the other and give other studio 2 ssd's test and see if size goes up?!
The best Mac Studio coversage - well done guys! Would love to see if you can get two SSD's to work in these machines.
@@m-stat9 Yes but now he’s using apple configurator 2 to reset the drives, probably will work
It will work with Apple configurator UNTIL they patch it for Apple employees only...
Seriously great content on a consistent basis - you guys are killing it. Articles across the net referring to your work taking the Studio apart! Deep dives and analysis on another level. Max Tech is the gold standard..
Hey dude you did so much effort look at 12:41 it is 6.04am.
Thank you.
Ayooooo I just saw that!
video idea, Swap the cooling systems to see if ultra can keep up with Max's Cooling system. Great work
Bad, BAD idea
I doubt it will boot lol
I had the same idea but i don't know if it would work
LOL you first
I could see LTT doing this. Would be interesting to watch for sure.
Awesome details! You got an even bigger advantage than any Tech TH-camr out there
Eagerly waiting for iPad Air 5 vs iPad Air 4 comparison.
7:00 It's somewhat clear that the M1 Ultra uses a vapor chamber for cooling, the difference isn't just in the cooler material
When you run thermal testing please mention your room temperature.
The room temperature is probably higher than the measured RAM temperatures. Normal room temp should be about 20-22 degrees Celsius.
Hey Max! You are absolutely killing it with these videos!! Hope you reach 1M subs soon, thanks for the video!
Or M1 subs
Could you hear the fans during these tests on either machine? I do audio work so it's important to me that the machine is as quiet as possible
No you never hear them and they never go over 1300rpm. Ever.
Hi Max have you tried the SSD from one Mac in the other and both in one machine to see if storage increases ?
Please run a simultaneous session of: 1. Xcode compile test (repeatedly over and over); 2. Logic w/ multiple tracks of white/pink/brown (different types of randomised) noise + plugins; 3. Video editor with a 4k edit. Time the compile tests. Track that time against the # of tracks in Logic as you increase them and the size and complexity of the edit. Artificial benchmarks are fine, but there is no real substitute for real-world workloads.
*Is there a high power mode in the Mac Studio like there is for the MacBook Pro? Maybe if you select it, it will enable both the CPU and GPU to run 100% at the same time. Try this on the MacBook pro aswell.*
Love your energy for these videos. Still waiting for mine to arrive. I got the M1 Max with 10-core CPU, 32-core GPU, 16-core Neural Engine
32GB unified memory
1TB SSD storage. Figured this should be more then enough power for me for hopefully years to come
Years to come? You wont be needing another computer DECADS to come unless apple decides to start slowing down their chips like they did in the old days lamo
@@samuelchimezie I still use a MBP 2012 fully upgraded and works great at a decade old. So this will will be a huge upgrade. I am thinking of getting the Studio display as well. Then see what comes out in the next year or so and get another MBP with M2 or whatever chip is avail then
the coverage on these products on your channel compared to even some of the biggest reviewers in the GAME is groundbreaking and it should be recognized fr
Maybe the reason why they can’t just idle one or two of the fans because other components like power supply need them on
I don’t know if you guys get enough credit for saying temperatures in Celsius. 90% of the world thanks you very much!
Honestly, I really had thought that the Studio was a bloated Mini with a big chip in it. Your tear-downs have completely eradicated that thought. I don't know if Apple appreciate you or not (probably not as you specialise in leaks!) but they should be appreciative of these Studio videos because they reveal the massive amount of effort and thought that Apple has put into creating these models. These units are so technologically dense that I think they are worth what is being charged and, by virtue of your videos, we now know that Apple could very significantly increase the performance of these machines any time they wish with a simple software update that allows another 20degC or so for the chip. Great job, guys!
Really impressed that you go to the Tear down path to show their differences. Keep it up and this will surely help potential buyers with their decision^ Hope for that 1M subs soon
The supposed full copper heatsink of Studio with the M1 Ultra is most likely using a vapour chamber from what you showed. And not surprising given the size of the M1 Ultra as compared to the M1 Max. But great in-depth look at the coolers and cooling capabilities.
Great review! Could you guys do a comparison between the Mac Studio base model and the MBP 16'" with the M1 Max chip?
7:01 the M1 ultra have a much larger heat pipe than that M1 Max. the heat pipe is not made of pure copper, infect they pulled almost all the air out of the pipe and added some water vapour in it, making it extremely efficient in transferring heat than copper by condensation. it is worth noting that this structure does not necessarily need to be done in a pipe looking way, in fact, it can be done in all different shapes as long as the condensed vapour can return to the chip efficiently so that it can be heated up again ( to transfer heat ), it will work. I expect the entire "Copper" instruction of that area covering the chip to have empty space ( low pressure when compared to 1 atm ) and contains some water vapour such that it will function like a heat pipe.
since heat pipe is much more conductive than pure copper can is capable to spread heat much faster, I can't think of any reason apple will not that the design described above.
Is it true that you can not upgrade the SSD?
Hm, maybe the gpu limit is not because of power and temp, but instead because cpu can’t control the GPU for the test at full speed while also being under full load on its own
Correct
It doesn't necessarily mean that it is Apple who limits the gpu performance when the cpu is maxed out. Its quite possible that something has to be prepared on cpu to provide the workload for the gpu. If cpu is maxed out, it becomes the bottleneck, and it doesn't provide enough stuff to max out the gpus.
I notice that in second SSD port at M1 MAX there are controllers that are missing in second port at M1 ULTRA which course on tests that second port could not be use for upgrade. Because there are chips in M1 MAX that would suggest possibility for upgrade. Can you please check that ?
*Max* is *maxing* out the M1 *Max* to check *max* power.
This is the maximum test ever
UPDATE to my original comment!! I Isolated out an issue that was causing the slow export times on both the Mac Studio and the Intel MacBook Pro. It turns out that if you add the "Outline" effect to a title the export becomes more than twice as long for some reason. So the Mac Studio with M1 Max is now twice as fast as the Intel MacBook Pro I was comparing it to. The M1 Max now exports the same 6:30 minute clip in 53 seconds and on the Intel MacBook Pro it now takes 1:10 minutes instead of 4:23 minutes. Super happy I found this out! My recommendation to anyone using Final Cut Pro is not to use the "outline" effect on a title if at all possible otherwise your export times will be much longer.
So far I am disappointed with H.264 performance on my Mac Studio with the M1 Max. In nearly every test that I've done the export times in Final Cut Pro are always about 36 seconds slower than my 2019 i9 15 inch Intel MacBook Pro. My expectations were super high for this but seeing as it is slower than older Intel machines when it comes to H.264 footage I may end up returning it until a fix comes to change this.
So your $2000 machine is 36-seconds slower than your $3000 machine while being completely silent and the fans don’t kick on if you connect two displays? Ehhh return away, I guess… someone else will snap up your folly of a refurb.
@@Wizzy0037 I actually have an update to this original comment, I forgot to edit it. I ended up Isolating out an issue that was causing the slow export times on both the Mac Studio and the Intel MacBook Pro. It turns out that if you add the "Outline" effect to a title the export becomes more than twice as long for some reason. So the Mac Studio with M1 Max is now twice as fast as the Intel MacBook Pro I was comparing it to which is what my initial expectation was. The M1 Max now exports the same 6:30 minute clip in 53 seconds and on the Intel MacBook Pro it now takes 1:10 minutes instead of 4:23 minutes. Super happy I found this out! My recommendation would be to never use the "outline" effect on a title unless it is needed otherwise the export times will be much longer.
Best mac's coverage in TH-cam ❤️❤️❤️
Do you ever sleep? You have done so many videos on all aspects of these new computers, it's amazing! Thanks for all the effort as the videos answer a lot of questions.
Thank you! This is the exact video that I wanted to know.
Seeing these results, I wonder if the chips are still operating as if they were inside something other than a Mac Studio? Thankfully, this is certainly something that can be corrected with software updates as the hardware design looks solid and there’s ample power and cooling on hand to crank performance higher.
1:45 In Luke Miani's trying to upgrade video, you could see that the SECOND SSD port had a BLANK PCB under the SSD port.. but here that isn't the case? Wonder if YOU try to upgrade the same way, will it work...
I dunno, the $1,999 one with 32-cores seems like the best bang for buck but this video helped me understand (on a hardware level) why that price isn't as crazy a jump. Neat video dudes, you guys were the first to teardown the studio I think.
It would be nice if you can transfer the copper heat sink to the M1 Max and see if there’s a performance increase.
The fans are not idle because they have to naturally cool the new power supply so it’s not all about SOC cooling
aluminum density 2.7 g/cm^3, copper density 8.9 g/cm^3, if the material used in both is the same, the copper will be around 3 times more weight.
The Battle of the Teardowns. Max vs Luke 🙂
I love it. Both great accomplishes and combined they give us Mac fans a superb insight. Thanks guys.
Tech is maxed out at Max Tech 🔥
Got a question, do softwares and programs need to optimise for m1 ultra over the m1 max?
Will apps need to update to use the extra power that comes with the chip?
Generally speaking yes. FCPX will need to update and probably countless other developers as well.
I’d say so also. Will be interesting to see the performance on these in the next year
maybe the ultra had a large vapor chamber instead of heat pipes on top...?
6 independent thunderbolt channel? You know what would be a nice speed test? Stripping 4 ssd in Raid 0 over 4 TB in the back of the Mac Studio.
I think that watt measurement is not a good indicator on the GPU usage. At a given frequency, watts indicate how many internal circuits are used. If a software does not use all functions of GPU because it doesn’t need , the watt will not be maxed but the usage will be at its maximum. The clock speed is to my point of view a good indicator of a potential throttle.
valuable information! thank YOU. I got the Ultra and I'm happy. :)
Should do a test after a month or two with high usage. Test it then to see if the cooling and functions still performs as it is when new.
You can’t tell if the aluminum is doing as well as copper as it’s cooling a different chip. To really test the difference you’d have to toss the copper heatsink into the aluminum system and Vice versus. Then test the differences.
Please cover a test like this for M2 Max vs M2 Ultra.. :)
This is a good good deal! I will get this spec for sure
I only speak for myself, and I know you guys are getting re-blogged by the verge and all the other big sites... so amazing on that, but as a general user, I want speed tests to help with my purchase. I want performance tests on different types of footage. I want what it's like to be in a timeline. The tear down is a whole other level of user, that I think is only a small portion of the creative users looking to buy this machine. All of this is to say that I'm eagerly awaiting the performance tests of the max vs the ultra. And PLEASE, every TH-cam channels is testing only in FCP, let's see Premiere and After Effects stuff too please! Love the channel, keep up the great work. Been telling everyone to stay tuned to your channel for these numbers. Been with max since the camera review days.
Do thermal tests with the Mac out in direct sunlight! Will have to wait for the summer though to really put it to the test.
I'm wondering after you replaced apples thermal paste with your own, did the temperatures improve?
Lol. As in you think this guy has better paste than Apple?
@@codycast Apple has been known to apply too much paste in the past.
@@codycast well actually Apple’s thermal paste in Intel macs used to be pretty horrible and replacing it did improve the temps
Hey! i've been hearing rumours about a constant hum fan noise at idle state, on the M1MAX, but not on the ultra. This video doesn't seem to agree with it, but what do we know about it ? seems that people are complaining out there, I'm planning to buy the M1 MAX Mac Studio for Music Creation. I am very sensitive to noise, so a fan noise problem would kill me literally - thanks in advance if you read my comment !
Sounds like they are putting it someplace like a metal shelf, or a hollow plywood desk that would act as a drum head for the vibrations of the motor. Stick to real wood, and you won't regret it. Particle board and mdf is for cupboards, and not even good at that. As for metal, sit on an empty oil drum and fart. You'll hear it further than a lion's roar. Kids bang on metal pans for a reason. Acoustics, never forget acoustics when setting up your PC layout.
@@arudanel5542 many thanks man for the precise answer
Been waiting on your recommendations for which Studio to get
Thanks for this video Max, incredible work! Please could you comment on the noise of the Max vs the Ultra - people on forums (and also TH-camr ArtIsRight) are complaining that the Max is loud at idle (louder than the Ultra). Please could you put this issue to rest - are they the same?!
Yes! Please!!!
I’ll take a listen. The fans seem to be identical and RPMs too.
@@MaxTechOfficial Thank Max Tech, some people complain about the nature of the sound (on reddit and MacRumors), the Max seem to resonate more than the Ultra and add some unpleasant frequency at the same 1325 rpm speed. For audio mixing and recording, that can be a bummer!
@@MaxTechOfficial Thank you!
Any news about this?
You guys are my go-to for all my Mac tech buys and reviews! Excellent job!!!
This is interesting, in Luke Miani’s video today, the second socket was bare and had none of the controllers like this one shows! I wonder why?
A bit late commenting on that, but Apple isn't limiting the GPU performance in that last CPU+GPU stress test scenario. Graphics pipeline actually needs some CPU resourses for draw calls for the GPU to do its job, and if you max out the CPU resourses with a 100% stress test, it surely starts bottlenecking the GPU performance. Nothing to do with "wattage" limitations or any other limitations for that matter. Hope that clears it out.
Thank you for the testing! And man you egot up super early..testing at 6am..or..havent slept yet?
Hi Max, what are you waiting before dropping a real life test of logic, final cut, etc.?
I tried to break Final Cut Pro and I am sure if it ran on windows vista at the time it would crash or blue screen but it never crashed even on a hot i7 MacBook Pro from 08 . Apple is the gold standard of tech
Could you also test the m1 studio after the new 12.3 update?
I noticed you said it will be completely silent is that accurate?
I’m an audio engineer and I often track vocals in the control room
With my 2018MBP I hear fan noise
I would consider getting a Mac Studio strictly off the fact that it is indeed silent if it actually is!!🙏🏻🙏🏻
GREAT Video... Thank you! I'm looking for an Adobe after effects ram preview and export comparison between the studio M1 MAX vs. the M1 ULTRA. I'm sure that video would get a nice chunk of traffic.
Excellent video. This is exactly the information I was waiting for to see if the Max Studio ran quiet when under stress like the M1 Mini does. Thank you :)
I am just waiting for some Lightroom performance tests. :)
Imagine how cool it would be if Apple put in a mini liquid cooler loop to compliment the heat sink.
Will the better cooling perform better on m1 max ?
You will be hitting 1M subs this year for sure.
Nice results.
Nice informative video. Wonder if the GPU is being throttled when the CPU is at full power, due to limitation of target power draw of the M1 Max SoC. Similar to how the PS5's SoC architecture works with AMD SmartShift
Honestly I don't know which Mac Studio to buy. I sold my iMac 27" i9 from 2020 for this haha. I do video and photo editing (no 3D stuff) with sometimes quite a lot layers. What do you think?
Amazing review as always. 1M subs coming up right away!
Great stuff guys. When are you going to try swapping an SSD from one Studio to another? and also trying to add an extra SSD from one Studio to another?
btw I think you account has been spoofed as I just received a message from someone pretending to be you, asking me to message them on Telegram, and offering me an iphone and watch for free...
run 3d and cinebench at the same time ! running 1 test at a time does not stress the cooler to it's max !
Watch out for those exposed caps. Big boom booms just like the ole' iMacs
8:45 working on a youtube video before 6am is pure dedication
The power supply is 370W with the Mac studio. Both the Ultra and Max don't use nearly as much power. I wonder if Apple gave the Mac Studio that power supply for future models going forward?
Or maybe future high power mode setting in future Mac OS?
High Power Mode already came out.
@@goldmastersimulations Not for Studio, it's not available in System Preferences yet.
This Apple choice of not ramping up the fans is weird for a pro machine that supposed to be designed to performance all the way. If you wanna keep at 60°C you ramp up the fan at the max. When the fan is at top speed, then you start to throttle the chip performance.
I was wondering if there is any possibility that this machine isn’t on “Performance Mode” and that is why it’s behaving like that?
Hey Max. How about a test using the M1 Max doing a 3 stream 4k multi cam in PP Thanks
Great breakdown as usual. Just want to point out that in the combined test, you might be seeing some CPU limitation of the GPU because the CPU is already maxed out by the Cinebench app. I noticed that when you ran Wildlife Extreme by itself, there was about 24% CPU utilization, so it seems like that test does require some CPU to fully push the GPU to its limits.
Seems like the studios need an update t allow more power over the Macbook pros
Hey Max, Great highlight of the Thermal throttling with CPU + GPU peged, I'm wondering since you've TGPro, please do test running benchmarck with TGPro forcing the fans to the Max (yes that'll be loud but that's for science and people that don't care about noise)
Very interesting, it's possible exchange aluminum heatsink with copper heatsink, and overclock M1 max?
I’m probably watching more than 80% of all the content about Apple Silicon on the internet, and let me tell you guys are on another level !
So glad to learn the Studio still has a dual fan with M1M, I wonder how it would compete vs a MacBook Pro with same specs…
*So this is the 24 core GPU version of the M1 Max?*
yes
Thanks for a great video. Let's see a Logic Pro X test
So I'm confused. It's not thermal throttling but still throttling? I've been leaning toward the Studio for the issue of not thermal throttling but if still throttling seems I might as well go for the Macbook??? Purpose: music production but I could really use the portability.
that M1 Ultra heatsink might be using a vapor chamber
Thank you again for all the hard work and doing these... Amazing. Every Mac user should watch... fascinating.
Once again this video confirms that you are providing the BEST reviews of the new Mac Studio in the internet! This time, I am confident you will reach 1M fans very soon!
Thanks for this review since this is the model I ordered because of your recommendation. Keep up the great job!