First review about sound... fan noise

[Confused-User]

As someone who never owned a Mini:

Is the 14” MBP M4 Pro (Base model) more capable when it comes to thermals than the M4 Pro Mini?

This is an ill-defined question. What does "more capable" mean to you?

If you mean "looking only at performance, which is superior?", then I don't think we have a definitive answer to that yet. It gets more complex when you factor in high performance mode, and the option of using fan control software.

But especially in this thread, there's been a lot of concern about noise. If you mean to ask "which has superior performance, while being effectively silent" or "...while being quiet enough not to be annoying" then you're in subjective territory, or you need to be very careful about defining terms and testing... well we don't have that answer either. Factor in fan control software and the question expands again.

There has been some suggestion that the M4 Pro 14" can run silent at higher loads than the Mini. That may be true, but I don't know for sure and I don't think we have enough test cases to be confident about that.

BTW, it's plausible that the 14" can run silent at higher load than the Mini, but that the Mini can run a higher absolute load if you don't mind the fan noise. If so, which is "more capable"? It depends on what you care about.

 

[Minga_]

This is an ill-defined question. What does "more capable" mean to you?

If you mean "looking only at performance, which is superior?", then I don't think we have a definitive answer to that yet. It gets more complex when you factor in high performance mode, and the option of using fan control software.

But especially in this thread, there's been a lot of concern about noise. If you mean to ask "which has superior performance, while being effectively silent" or "...while being quiet enough not to be annoying" then you're in subjective territory, or you need to be very careful about defining terms and testing... well we don't have that answer either. Factor in fan control software and the question expands again.

There has been some suggestion that the M4 Pro 14" can run silent at higher loads than the Mini. That may be true, but I don't know for sure and I don't think we have enough test cases to be confident about that.

BTW, it's plausible that the 14" can run silent at higher load than the Mini, but that the Mini can run a higher absolute load if you don't mind the fan noise. If so, which is "more capable"? It depends on what you care about.mean which machine is going to be more silent under the same load.

Sorry, I meant which machine is going to be more silent under the same load?

 

[halo9]

A proper stress test for a specced out M4 Pro Mini, to test for thermals and fan noise:

Hint: It passes with flying colours.

Very interesting, especially at the end when he runs everything at once. Amazing what these machines can do. If only he included a recording of the machine noise when the fan was running and a db reading before and during, would have been a perfect demonstration. Guess youtube does have some interesting videos 😝

 

[yalej]

The real stress test is 4K gaming. I tried for a little bit with an (old) game I play called EVE Online (which is not exactly known for being graphically demanding). 4K resolution, max graphics settings, and FSR 1.0 (all that the game currently supports) set to Balanced. CPU (14/20 Pro) temp reached 107C and fan noise was noticeable, I measured 45 db about 2-2.5 feet away from the mini (measured with Apple Watch, so not exactly the most accurate I am guessing). Fan had max rpm of 3100. Interestingly, GPU usage was only 80%. FPS were 65-80.

However, this peak temp was not permanent, and it went down to about 95-99C after maybe 20-30 seconds even while nothing else was changing on the screen and I wasn't doing anything. Not sure why. Along with that, the fan noise goes down and all-but disappears.

 

[ubinko]

Another reviewer who's never heard the fan on the M4 Pro Mini.

I feel at this point, we can conclude that the fan noise is not a serious issue (for most use cases).

 

[ubinko]

Huge stress test for M4 Pro Mini. Incredible machine. It barely got warm and the Fans hardly kicked in.

 

[Confused-User]

Sorry, I meant which machine is going to be more silent under the same load?

That's a much better question, and I don't think we have a definitive answer yet.

Once you have the answer, though, you're not done. Since you can set the max fan speed, you can specify any limit you like on fan noise. And if you do that, which machine does better? The answer may be the same as the first question, but it may not be! That's because Apple's fan curve may not be optimal for you, and it may not be a curve, but rather have set points, in which case your particular load profile may fall at a bad spot. So you would have to test this yourself to really know.

As yes, it's not great that it's hard to figure this out. If you're feeling lazy, I guess you can wait until the Studio M4 ships...

 

[Confused-User]

Another reviewer who's never heard the fan on the M4 Pro Mini. [...]
I feel at this point, we can conclude that the fan noise is not a serious issue (for most use cases).

Huge stress test for M4 Pro Mini. Incredible machine. It barely got warm and the Fans hardly kicked in.

If you have a few moments, I'm sure many of us who do not spend a lot of time watching youtube would appreciate a summary.

 

[nathansz]

Can you confirm Macs Fan Control works with M4? I heard that it didn't work properly on Sequoia (and Sonoma starting in 14.7):

https://forums.macrumors.com/threads/macos-14-7-not-letting-manually-control-fans-speed.2437236/

Mac M3 Pro (Apple Chip) updated with macOS Sequoia 15.0 App not working now · Issue #776 · crystalidea/macs-fan-control

I was using Macs Fan Control for months without an issue. Yesterday I updated the operating system to Sequoia 15.0 and now the app doesn't turn on the fans at any configuration. Not with custom spe...

github.com

Based on reports of it not working on system even after downgrading, I suspect the firmware updates bundled with the newer OS break Macs Fan Control.

just noticed something that I thought I should share here

Macs fan control cannot turn the fans on if they are not already on

once the fans are on, they can control the speed

 

[nathansz]

The real stress test is 4K gaming. I tried for a little bit with an (old) game I play called EVE Online (which is not exactly known for being graphically demanding). 4K resolution, max graphics settings, and FSR 1.0 (all that the game currently supports) set to Balanced. CPU (14/20 Pro) temp reached 107C and fan noise was noticeable, I measured 45 db about 2-2.5 feet away from the mini (measured with Apple Watch, so not exactly the most accurate I am guessing). Fan had max rpm of 3100. Interestingly, GPU usage was only 80%. FPS were 65-80.

However, this peak temp was not permanent, and it went down to about 95-99C after maybe 20-30 seconds even while nothing else was changing on the screen and I wasn't doing anything. Not sure why. Along with that, the fan noise goes down and all-but disappears.

if you REALLY want to stress test try nhl legacy in RPCS3

I bet it will keep the fans going full blast

 

[DankeBrutus]

A proper stress test for a specced out M4 Pro Mini, to test for thermals and fan noise:

Hint: It passes with flying colours.

I watched this yesterday and it has the same problem a lot of videos and articles out there have talking about Mac Mini performance - no concrete numbers.

 

[DankeBrutus]

there's lots of speculation going on in this thread...

I have a Mac mini M1, M2 and M4 (non-Pro). I ran a HandBrake re-encoding job (H.264 -> H.265, using the CPUs, not VideoToolbox). this, of course, maxes out all available CPU cores.

M4 P-core max. possible/allowed clockspeed: 4464 MHz
M2 P-core max. possible/allowed clockspeed: 3504 MHz

M4 fan at idle: 1000 RPM (inaudible)
M2 fan at idle: 1700 RPM (inaudible)

both, M2 and M4, consume approx. 24 W with all CPU cores maxed out.

the fan in the Mac mini M2 stays at 1700 RPM, all the time.
the CPU clock speed stays at 3504 MHz during re-encoding, all the time.

the fan in the Mac mini M4 ramps up to approx 2100 RPM (clearly audible, mildly annoying).
after 10-15 minutes under full load, the CPUs start to throttle (worst seen approx. 3300 MHz, settles after a while to approx. 3800 MHz).

so... with M2 and M4 both using roughly the same amount of power (~24 W) when running the same job, I can only draw one conclusion from this: the cooling solution in the Mac mini M2 works better, period.

What temperatures were the M4 and M2 chips running at during the job? How long did either take to complete?

 

[DankeBrutus]

Fan had max rpm of 3100.

Do you mean that the maximum RPM of the fan is 3100 or that was just the peak you saw?

 

[ravenofdoom]

I’m not sure why it’s so hard to get these concrete numbers from reviewers — nobody on YouTube is doing a simple test of having a taxing 4k game (or a benchmark that runs a game on loop) measure 3 parameters on both the m4 and m4 pro and log them over time
1) internal temp,
2) fan rpm,
3) db measured 2 feet away (with a control measurement of room floor dB beforehand).

Yalej did this for the m4pro. I suspect 45db is the peak fan noise on this unit (excluding max performance mode) if the room was close to silent / 0-15db. I’d like to see how that # compares in a like-for-like test with the m4 mini.

 

[DankeBrutus]

I’m not sure why it’s so hard to get these concrete numbers from reviewers — nobody on YouTube is doing a simple test of having a taxing 4k game (or a benchmark that runs a game on loop) measure 3 parameters on both the m4 and m4 pro and log them over time
1) internal temp,
2) fan rpm,
3) db measured 2 feet away (with a control measurement of room floor dB beforehand).

Yalej did this for the m4pro. I suspect 45db is the peak fan noise on this unit (excluding max performance mode) if the room was close to silent / 0-15db. I’d like to see how that # compares in a like-for-like test with the m4 mini.

A modern game benchmark that runs natively on M-series SOCs will of course hit both GPU and CPU but load will be mostly leaning on the GPU. I think the handbrake test done by @mikeboss is good because it is a task that will max out CPU usage. Especially with those power cores. We just need granular metrics to see how the CPU handles it. Like what temperature was the CPU throttling at? What was the measurable impact on performance? An easy measurement to bring up would be how long the M4 took to finish the handbrake job. If the M4 Pro throttles at like 105˚C but throttles down to a more manageable 85˚C and still beats out the M2 Pro with the more beefy cooling solution then the M4 Pro would be suited to the chassis.

It would also be helpful to see how we can adjust things. I checked earlier and, at least according to their website, TG Pro is not yet fully compatible with M4. The highest RPM that @yalej found was 3100. But if the fan's RPM maxes out at, say, 5000-7000 RPM then the user could make make adjustments to the fan curve to keep the SOC in check.

edit: I recalled a Chinese review being posted to the Apple subreddit that contained some numbers.

Here it is starting at 17:11

So I don't know Mandarin and had to rely on the Translate app on my phone for some bits of the video but that timestamp is pretty well all English and numbers. In, I believe, Cinebench R23 while in high power mode for the non-binned 14/20 core M4 Pro, 48GB memory, we have a temperature of 82˚C, a fan speed of ~4900 RPM, and the SOC was not throttling. Now the temperature could perhaps be called into question as I can't tell the exact sensor the reading comes from but benefit of the doubt let's say it's an average. That's pretty good in my opinion. My desktop with a Ryzen 5 3600 and a Thermalright Peerless Assassin 120 will run at 68˚C at it's highest during the handbrake jobs I've been throwing at it over the past month (H.265 10-bit, slowest preset, RF23)

 

[yalej]

Do you mean that the maximum RPM of the fan is 3100 or that was just the peak you saw?

Peak I saw.

I think also one thing to keep in mind is the ambient temperature. I'm trying to get the fan speed I saw previously and having trouble getting it above 2300 rpm (where it is pretty quiet) even when the CPU temp > 103C. I want to say the ambient temp is maybe 70-72F in the room I'm in, but I don't have a thermometer in this room (house set to 70F downstairs).

 

[nathansz]

Do you mean that the maximum RPM of the fan is 3100 or that was just the peak you saw?

16” MacBook fans go to 5700 max fwiw

 

[halo9]

halo9 said:
So without speculation the new minis do have more powerful chips that put out more heat faster due to their processing power, but they do it for less time, so less overall

But in general, it won't be true, because Apple decided that they could afford to give up some efficiency in favor of higher maximum performance. So the M4 will often use more total energy for a specific fixed task than will the M3.

You are right. I was trying to simplify it but ended up too vague. My point was the M4 chips draw more power so inevitably they will put out more heat. They should finish the task quicker though, so that heat is output over a shorter period of time. This requires a more responsive cooler and aggressive fan curve. Hence the fans are noticed more.

In the encoding task I referenced, even though the M4 pro drew a higher current during the task, as it finished quicker than the M2 pro, it’s overall efficiency was better. Performance per watt that is.

My shipping estimate is 6-12 December now so still have to wait a while to test myself.

 

[Confused-User]

You are right. I was trying to simplify it but ended up too vague. My point was the M4 chips draw more power so inevitably they will put out more heat. They should finish the task quicker though, so that heat is output over a shorter period of time. This requires a more responsive cooler and aggressive fan curve. Hence the fans are noticed more.

In the encoding task I referenced, even though the M4 pro drew a higher current during the task, as it finished quicker than the M2 pro, it’s overall efficiency was better. Performance per watt that is.

It's not clear from your response if you understood that in fact performance per watt with the M4 will often be worse, not because the M4 is inherently less efficient (it's inherently MORE efficient) but because Apple chose to allow it to move further out on the perf/power curve. If you did, sorry for belaboring that point.

The interesting thing here is that you could choose to limit fan speed to a level that's quiet enough for you, and... then what? Is the resulting somewhat-limited M4 Pro Mini better that the M2 Pro Mini, or will the M2 Pro in the old chassis perform better because it can use more power without fan noise? We don't even know if that latter part is true - can it use more power without making the fan noisy? While lots of other questions have been starting to get answers, as far as I know this hasn't yet.

There was one claim that both the M2 and M4 Pros would max CPU at 24W, and that that would cause the M4 to be noisy while the M2 was not. That doesn't quite answer my question, though it starts to get close, but I'd like to see confirmation of that.

 

[SkweeBop]

Here's the base model review

 

[hgfjhbvytjdnb]

Taking any of the YouTube "influenzas" input into anything that goes beyond how to make a-thumbnail-with-the-silliest-face-without-looking-like-you're-not-well is not the best idea. I'll wait for real user feedback from large sample of users before I decide whether to pull the trigger or not.

 

[hise]

I bit the bullet and pulled the trigger on an M4 Pro 48/512 today. The M2 Max Studio is going back on return. I will accept the possible noise pollution from the M4 Pro until the M4 Max is available as the cost of a short-term rental. I'm not sure whether this is the best decision, but there comes a point where decisions must be made, and you need to get back to work and play. Realistically, we are blessed with M series processors that can handle almost anything within reason, and paralysis by analysis is real.

 

[Confused-User]

I bit the bullet and pulled the trigger on an M4 Pro 48/512 today. The M2 Max Studio is going back on return. I will accept the possible noise pollution from the M4 Pro until the M4 Max is available as the cost of a short-term rental. I'm not sure whether this is the best decision, but there comes a point where decisions must be made, and you need to get back to work and play. Realistically, we are blessed with M series processors that can handle almost anything within reason, and paralysis by analysis is real.

I think this is a very smart decision.

If you have time, I (and, I'm sure, others) would love to hear any experiences you can share (preferably with quantitative info, but even without) about noise and fan behavior. Especially, what kind of performance you can get while keeping the fan quiet, and what that fan speed and performance is.

 

[AAPLGeek]

If you're a Topaz VAI user, expect some fan noise

 

[ravenofdoom]

I had a brief exchange with the guy in those two m4 and m4pro videos. He confirmed what I wanted to know — that at no point did the fans ever engage in the m4 to being audible, even in the most taxing cpu+gpu tests performed on the m4pro (where he noted it was audible). I think I have what I came here for to make the right choice for a silent machine that will sit all day feet away from me in my living room. When you compare the speeds (geekbench), the non GPU performance of the base m4 is on par with the m2pro, and the m4pro is only about 1.5x faster CPU and 1.8x GPU. Coming from a 7th gen i5 intel nuc, I’m still getting a 5x bump in cpu and GPU and so I’ll take the silent machine.

	M4 mini	MBP m2pro	M4 mini vs MBP m2pro	M4pro mini	M4 vs m4pro mini
CPU single	3767	2676	0.7103796124	3925	1.041943191
CPU multi	14677	14843	1.011310213	22669	1.544525448
GPU (open cl)	37773	53472	1.415614328	69867	1.849654515