If there’s one thing I hate more than overpriced goods and services, it’s unwanted noise from electronics and computers. Coil whine and fan noise are public enemies #1 and #2 when it comes to choosing or building a computer system.
Today, we’re going to take an in-depth look at how silent Apple’s M2 Pro Mac mini really is. We’ll look at a variety of measurements and tests that I performed, and I’ll give my verdict on whether the M2 Pro Mac mini is a silent beast or a noisy monstrosity.
First, let’s talk briefly about the 2 types of noise I’m going to examine. The first is a type of electrical noise, also known as coil whine, which occurs when an electric component vibrates as power runs through it. This can range from a low buzzing sound to a high-pitched whine, and can be really annoying. If you have a gaming PC, it’s not uncommon to hear high-pitched coil whine from your GPU when running at high frame rates, such in game menus.
Coil whine is also the worst type of unwanted computer noise, as there’s not much you can do about it if you get it, other than to return the computer or component and exchange for something else.
When it comes to the M2 Pro Mac mini, at idle, there is a very slight buzzing sound that can be heard if I put my ear all the way up to the machine. However, I can’t hear it at my normal sitting distance of about 2.5 feet away. I’m also happy to say that in all the tests I ran, there was no obvious increase in coil whine when the system was under load.
I even booted up Divinity: Original Sin to the main menu, which would always get the GPU in my desktop computer whining like crazy. On the M2 Pro Mac mini, there was only a very slight increase in the coil whine buzz, but again, I had to put my ear closer to the machine to hear it. So coil whine is not really an issue with the M2 Pro Mac mini.
The other type of unwanted computer noise is fan noise, which is generated by cooling fans that are trying to remove heat to keep the system cool. Usually, the hotter a computer runs, the faster the fans need to spin, and thus the more fan noise generated. Having high-quality fans and also smart fan profiles that ramp up fan speeds just high enough to effectively cool the system are essential to reducing the impact of fan noise.
Since we are talking about the Mac mini today, and Apple doesn’t make any water-cooled computers, I won’t go into detail about a related source of noise from the cooling system, pump noise, which is generated by water pumps in water-cooled computer systems.
Noisy Mac Studio
When Apple released the M1 Max and Ultra Mac Studio, many people had reported that it was unusually loud compared to other Macs. There were complaints both about electrical whine and fan noise. And tests showed that it indeed emitted more noise than almost every other Mac at idle.
I had one myself, and while my unit didn’t emit as much coil whine as some others reported, I did discover that the default idle fan speed of the Mac Studio was set to 1300 rpm and placed about 3 feet or 1 meter from me on my desk, was clearly audible. Shortly after release, though, tools like Macs Fan Control were released and allowed you to lower the fan speed down to 1100 rpm, which made it nearly inaudible at a normal sitting distance.
With how cool the Mac Studio was running, Apple seemed to make a mistake by setting too high of a default fan speed at idle. Let’s take a look and see if Apple made a similar mistake with the M2 Pro Mac mini.
First, let’s take a look at the noise levels specified by Apple. Do note that any decibel numbers should only be used for relative comparisons only. It’s hard to know exactly how these numbers were measured and the environment they tested is likely very different from where you will actually be using the product. But we can assume they tested things with similar setups, so the relative comparisons are still useful.
Apple claims the M1 Mac mini at both idle and wireless web usage emits 5 decibels of noise. However, we need the Mac Studio specifications say it emits 15 decibels, which is in line with the reports that it was indeed much louder. Thankfully, the 2023 Mac mini with M2 and M2 Pro chips were tested to have the same result as the M1 Mac mini at 5 decibels of noise. That suggests Apple did not make the cooling system bottom out at too high a level as they did for the Mac Studio. So far so good!
Idle Temps and Fan Noise
Let’s take a look at what temperatures and fan speeds look like at idle. Shortly after booting up my computer, we see that CPU temps were around 33°C. The M2 Pro Mac mini runs it’s single fan at a minimum of around 1700 rpm while idle. At this speed, I can only hear it if I get close to the machine and press my ear up to it. I can’t hear it at all from my normal sitting distance of 2.5 feet away. So at idle, the Mac mini is silent.
100% CPU Temps and Fan Noise
Now let’s jump to stressing the computer out. If putting it under heavy load isn’t able to ramp up the fan speed and noise levels, then there’s really no use doing intermediate testing with other applications. Used a program called Endurance: CPU Stress Test to put the computer under 100% CPU load for about 30 minutes.
After using the computer for a while, the CPU started the test at around 40°C, and immediately started increasing. After about 2 minutes, the performance cores were already in the 80 degree range, but the fan speed hasn’t ramped up at all yet. As we approach 3 minutes, some CPU cores pass 90 degrees, but still no fan speed changes. It isn’t until we start to see multiple CPU cores in the high 90’s around the 5 minute mark that the fan speed starts to gradually ramp up.
The fan speeds cross 1800 rpm, and then 1900 rpm at around the 6 minute 30 mark. The system continues to heat up more and more and we see fan speeds reach 2000 rpm as we approach 10 minutes. It takes another 5 and a half minutes for the fan speeds to reach 2100 rpm. Ultimately the fan speed did get in the 2200 range, but seemed to hover there as the test ran into the 20 to 30 minute territory. Several times, we can see the CPU cores reach 100°C.
It’s clear that Apple has put a fairly relaxed fan curve on the Mac mini. Although the fans can be turned up all the way to 5000 rpm, we’re only seeing the range of 1700 to 2200 rpm being utilized, at least when we are just using the CPU.
In terms of noise levels, the machine does become audible from my sitting distance at 2200 rpm, but the sound level is very low and just a soft whirl. It’s not what I would call distracting or annoying, so if I needed to deal with extended 100% CPU usage for a long time, it’s not a problem.
100% GPU Temps and Fan Noise
Now let’s see what happens when we have 100% GPU usage. To run this test, I used Unigine Heaven 4.0. For the first 9 minutes or so of this test, we see GPU and CPU temps steadily climbing, but the fan speed remains at the base 1700 rpm level. While this is a GPU stress test, there is some CPU being used, and the system on a chip design of Apple Silicon probably means there some correlation between CPU and GPU temps.
At around 10 minutes, the fan speed has ramped up to 1800 rpm, even though none of the temperatures have exceeded 90°C. This means the fan curve works differently for GPU usage and CPU usage, and Apple seems to be more aggressive at using the fans to keep temperatures down when the GPU is being used.
Indeed, fan speeds continue to ramp up, reaching 2000 rpm around 12 minutes, but we can see that overall temperatures have actually fallen a bit, with the CPU average at 81 (down from 85 previously) and the GPU cluster at 84 (down from 88). This trend continues as fan speeds reach 2200 rpm again around 14 minutes, while CPU temps have fallen to 79 and GPU to 82.
The fan speed seems to stop ramping after breaking 2300 rpm at around 15 minutes and getting as high as 2350 rpm. It then hovers between 2200 and 2300 for the remaining time, while CPU temps are down to 75 and GPU temps are also below 80.
So we can conclude that GPU utilization causes the fan speed to ramp up faster and reach a higher plateau than CPU utilization. The noise signature between the CPU and GPU tests are not that different at all. In both cases, the max rpm of around 2200 to 2300 only results in a slightly audible soft whirl noise that is not annoying.
100% CPU and GPU Temps and Fan Noise
For the final stress test, I’ll max out both CPU and GPU utilizing by using both of the tools I ran previously at the same time. This time, we see that the Mac mini’s thermal limits are really starting to be pressed.
I should start by noting that while I was setting up the test to be run and still getting my screen set up, temperatures shot up really quickly with the CPU exceeding 100 degrees, reaching up to 107 at one point, and the GPU cluster in the high 80’s. Correspondingly, fan speeds also shot through the roof and quickly ramped up past 4000 rpm. I think the system got freaked out and wasn’t ready for the high temps so quickly, which is why the fans lit up like a rocket engine, and sounded like one too.
Once I got my windows in place, however, and started the CPU test over again, it never reached that high anymore. Starting from the 2300 rpm of only using GPU, adding on the 100% CPU load did cause a quick ramp up of fan speeds to 3000 rpm. From there, it seemed gradually rose all the way until the 3300-3400 rpm range, where it seemed to stay. It never approached 4000 rpm again, though.
At 3300 rpm, the Mac mini is sufficiently loud to where I would not want to be using it at that speed.
Fan Noise Sound Samples
So we’ve seen what happens when we stress the CPU, CPU, and both. Stressing only the CPU or only the GPU, the fan speeds don’t go past the 2200-2300 range, and the soft whirl noise is acceptable to me even for some prolonged use. Stressing both CPU and GPU causes fan speeds to ramp up to the 3300-3400 rpm range, which is too loud to be used regularly.
Next, I want to give you a sense of what the Mac mini’s fan actually sounds like. To do this, I’ll record the fan noise while ramping up the fan speed. Keep in mind that it’s hard to know the actual noise level in this comparison, since I have to adjust the volume of the sound and it also depends on your own volume settings. But it should give you an idea of the relative noise levels at different fan speeds, and what that noise sounds like. Watch the video to hear the sound samples.
What About Actual Usage?
So what do these findings actually mean? Is the Mac mini silent or not for actual usage? Remember that the behavior of the Mac mini fan is such that it doesn’t ramp up from the minimum 1700 rpm until several minutes of prolonged 100% usage had occurred. That means for most daily productivity tasks, such as browsing the web, working on documents, and even some computing intensive tasks in bursts, you shouldn’t expect the fan to ramp up at all, and the Mac mini will be completely silent. And indeed, that’s been my experience.
But what about those prolonged, intensive productivity tasks? Well, it will depend on what you’re doing. If you’re running something that will put intense loads on the CPU and GPU for an extended period of time, then yes, you can expect the fans to ramp up. But often, even those tasks won’t result in 100% sustained load. For example, here’s me rendering a video in Davinci Resolve with the “optimize for speed” setting turned off, and we can see the GPU is heavily utilized, but sometimes it isn’t. Those pauses in utilization allow temperatures to stay low and the fan speed to remain at the minimum.
If you really do have prolonged, intensive tasks, you probably also wouldn’t be using the computer at the same time, but rather waiting for it to complete the task or going off to do something else instead. In that case, the fans being loud while working on that job aren’t going to be much of an issue.
The primary exception to that may be something like gaming. To test this, I ran Final Fantasy XIV with graphics settings on high, which put 100% load on the GPU. There was also some load on the CPU, and the results were similar to the GPU stress test. The fan speeds eventually stabilized around 2300-2400 rpm, with both CPU and GPU temps around 80°C. This means the fan noise was an audible, soft whirl, but if you’re playing a game, you probably also have music and sound effects, which will easily hide the fan noise. I think most games won’t put 100% load on both the CPU and GPU, so it shouldn’t cause the system to ramp up the fans to be too loud.
In summary, I’m impressed by the M2 Pro Mac mini’s ability to remain silent most of the time. Even when power is needed, the fans ramp up but aren’t too loud unless you have a rare use case that stresses both the CPU and GPU for prolonged periods. It’s certainly a lot better than comparably sized computers like some of the Intel NUC systems I’ve tried in the past. I conclude that the M2 Pro mac mini passes the test of being a silent computing beast.