The only wall here is that your argument is entirely fallacious
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
First review about sound... fan noise
- Thread starter Cape Dave
- Start date
- Sort by reaction score
The only wall here is that your argument is entirely fallacious
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):
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.
You are working from a fundamental misunderstanding. All form factors that Apple sells, excepting possibly the Mac Pro M2 Ultra, will cause higher-end chips (Maxes, and some Pros) to "throttle" (terrible word, as I've written before, but it's what we've got). In fact, as a reductio ad absurdum, even baseline M chips are throttled by their cases' design, since none of those cases have liquid cooling, and even base M chips will run faster if so cooled.
Obviously, liquid-cooling an M4 (or M1) is silly. The real-world performance benefit is minuscule in all reasonable scenarios.
So the interesting question isn't "does the M4 Pro mini throttle?". Of course it does, just like all the others. The interesting question is "how much more or less would it throttle in the M1 mini's case?", and by extension, "should Apple have continued to use that case (or some other larger one)?". The answer to that question is not known, and none of the discussion here has thrown any light on that question.
That question is totally separate from "Does the M4 Pro mini get too loud when running at maximum?" (apparently, yes for some loads and some people). It also says nothing about "Will the M4 Pro still have good performance if you force the fans to run quietly?" (about which we have no data as yet).
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):
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...
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.
I just used it on my MacBook Pro m4 pro
Works as expected
That's a faulty assumption, and not in a minor way, for two main reasons:
1) You're assuming that any increase in heat dissipated beyond what the M4 puts out will cause the fans to ramp up. But you don't know that. In fact that's incredibly unlikely. All you know is that the M4 can't push the fans enough to make them ramp. There is some (unknown) amount of headroom beyond the M4's maximum heat dissipation that the fan can handle before ramping up.
2) Even assuming for the moment that the M4Pro had to run slow enough to dissipate only as much heat as the M4, it would still have markedly better performance (though not nearly as good as it does in normal conditions). That's because each core would get only a fraction of the power the base M4 cores get, but because cores are more efficient the lower their power, each core can do more work per watt. In other words, the entire chip would run more efficiently. This is in fact the (successful) strategy Qualcomm has employed in its recent Oryon ARM chips to improve their multicore performance: use more cores, running at lower clocks, to improve overall efficiency.
As I explained above, you might not have to do that... but I think you've just identified an opportunity for an indie developer (no, not me): Make a fan speed tool that ramps down the fans when either:
1) There has been user interaction in the last 15 seconds (configurable), or
2) Certain software is in the foreground (Apple TV, VLC, other media players, etc.)
You're making a faulty assumption here. The process (N3E) is more efficient, but that's not what matters. What matters is only total heat dissipation from the two chips (M2 Pro and M4 Pro). Apple is clocking the M4 Pro much much higher than the M2 Pro, and there are more cores as well. So the M4 Pro dissipates a lot more heat than the M2.
Unless somebody benchmarks the internal temp of the box vs. fan speed / dB vs. computations performed, I think it’s equally a faulty assumption to say that the m4pro would have a markedly better performance if constrained to the inaudible fan speed of an m4. What that fraction of power each cores get against the net computations remains to be seen. Maybe it’s better by 10%. Maybe 50%. Who knows. My point was that to a consumer, if you want a universally quiet mac mini without fiddling with overriding fans and underutilizing / throttling down the cpu, the clear option seems to be the base m4 (unless you want Tb5 / faster SSD)
+1.
It’s pretty simple. Beefier heatsink and larger fan pull the heat away from the chips more efficiently and at lower RPMs.
It’s clear that thermals weren’t Apple’s main concern when redesigning the enclosure. Looks like they just took the size of the existing M2 logic board (5" at it's widest) and basically designed the smallest possible case around it. Even a slightly larger case, like 6”x2.5” instead of 5”x2” would’ve allowed a much larger, beefier cooling system needed for the Pro chips. But I guess that marketing was really going for a “wow” factor to create the smallest possible desktop machine here.
It almost seems like Apple has painted themselves into a corner with this new Mini redesign, especially for the Pro chips. What happens when there’s even more powerful M5 and M6 chips in the future? That puny heatsink would cause the fan to ramp up even more. The PSU would also draw more power as the chips get more powerful causing the limited thermal system to become less and less efficient.
Last edited:
This is the fundamental disconnect. You have learned a fact, but it's not the only fact. It's not even, by a lot, the most important fact.
Yes, size does matter, but not the way you think. What matters MUCH more is the cooling capacity, which is in turn constrained by size, but not at all in a linear way.
Imagine that I removed the top of the M4 Pro Mini completely. Now the volume of the case has become the volume of the room I'm in. Do you think that that would help cool it significantly? Of course not. It would likely make the situation worse.
Nearly all of the cooling capacity comes from air being moved across hot surfaces. There is a discussion to be had about heat sinks, and how well they're coupled to the case, and how well the case radiates, but heat sinks are irrelevant to steady-state work from the standpoint of thermal capacity (once they've soaked up all the heat they can, that's it), and neither the old nor new Mini cases can radiate away a lot of heat compared to the air cooling.
So if you want to know if the old or new Minis are better at cooling, what you really need to know is how well they move air across the hot surfaces (which are the heat pipe and/or radiator surfaces). Honestly, I have no idea how they stack up that way, but so far nobody else has had anything meaningful to say about that either.
So how is size even relevant to this discussion? Tricky to say because we seem to be talking about two different things at the same time: noise, and cooling capacity. They're not the same thing at all. What *I'm* talking about right now is what you were talking about - cooling capacity.
As you know, size matters because the larger a fan is, the more air it can move (at constant RPM). But that's not the only thing that matters, because good engineering (directing the airflow across radiating surfaces) is far more important. Really bad airflow can completely eliminate the benefit of a fan. Both ingredients are necessary.
It appears that the fan in the M4 Mini is roughly the same size as the fan in the M2 Mini, but I don't know that, and would welcome any hard data on that. The depth of the fan matters too, so what you really want to know is the CFM rating of both at constant RPM.
Now, switching to the other point, about sound: You could certainly build a case the size of the old Mini that definitively was quieter than the new case, if you used larger fans and a large radiator (what people are calling a "heat sink", though that misses the point here), and you carefully engineered the airflow. But that's not what everyone here has been talking about. They've been talking about the actual M2 Mini case, which did not have those advantages. So far, nobody has shown any facts about the old case that could prove it's better than the new case.
You want to argue that Apple should have build a new case for the Pro at the size of the old mini, with proper cooling, rather than the new case? Be my guest. It wouldn't be all that different from a Mac Studio, which might be why Apple made the choices they did.
Again, bad argument. The question isn't whether the Studio is better at cooling than the new mini - it obviously is. The question is whether the old Mini is better. You don't know, and so far nobody here does.
It's really not... for most values of "markedly better". I mean, is the M4 markedly better than the M3? You're looking at at least that much improvement, though I'm way too lazy to put up some plausible numbers.
M4 = 4P6E, M4P = 10P4E. Leaving the E cores out as being fairly insignificant (they are fewer, and use much less power), you're looking at running each P core of the Pro at only 40% of the power of the M4 P cores. (Yes, the uncore of the M4Pro will take some of that away, but not that much.) Does anyone have the performance/power curve for the M4 P cores handy?
Considering that that's 1-3 entire generations of CPU improvement, that's quite a lot.
Sure. But that leaves open a big space for people who need more grunt than the M4, who don't like noise. Will the M4 Pro Mini work for them? Open question, and the answer will depend on the specifics of their workloads.
As far as I'm aware, every processing unit designer has taken full advantage of smaller manufacturing methods by packing more transistors in the same or even larger die size as well as pushing the frequency limits — nullifying the efficiency gain from at least one perspective.
Apple M2 Pro chip, 32GB unified memory, 8TB SSD:
- Power Consumption:
— Idle: 7W, CPU Max: 100W
- Thermal Output:
— Idle: 24 BTU/h, CPU Max: 358 BTU/h
Apple M4 Pro chip, 64GB unified memory, 8TB SSD:
- Power Consumption:
— Idle: 5W, CPU Max: 140W
- Thermal Output:
— Idle: 17 BTU/h, CPU Max: 478 BTU/h
Mac mini power consumption and thermal output (BTU) information - Apple Support
Learn about the power consumption and thermal output of Mac mini computers.
By the way, this is where it began:
Apple M1 chip, 16GB unified memory, 2TB SSD:
- Power Consumption:
— Idle: 6.8W, CPU Max: 39W
- Thermal Output:
— Idle: 23.2 BTU/h, CPU Max: 133 BTU/h
P.S. Ever since the comments of changing the Mac mini (or any Mac) for the sake of something different to look at, I’ve been saying it’s small enough and what’s wrong with the spare cooling headroom?
P.P.S. I’m not concerned and still excited for my M4 Pro mini.
A quick look at some geekerwan numbers:
An M4 P core running at 4.47GHz scores 11.72 in SPECint, using 7.21W. 11.72/7.21 = 1.625 S/W
The same core at 4.04GHz scores 10.67, using 4.95W. 10.67/4.95 = 2.155 S/W
That's 33% increase in efficiency due to just 10% less power to the core. Expect further improvements as power drops further.
That means that I was likely understating how much faster the M4Pro would be compared to the M4, *at equal power levels*. It's likely to significantly exceed the most optimistic figure suggested by @ravenofdoom in his post (50%). That most definitely counts as "markedly better".
No coole whine, a little bit of whistling from the fan during high rpm.
Sorry, but I already returned the Mac mini.@maluba27
Extremely useful! Asking for a lot, but is there any chance you can place your phone as a dB meter maybe 2 ft away and snap a screenshot of your results pre-fan and post-fan (as in your video)?
Audio Spectrum Analyzer Pro
Audio Spectrum Analyzer is a powerful real-time audio analysis app. Full of features yet very simple to use. Sound level meter (db meter a-weighted, noise rating, noise criterion) included. Feature Highlights: • Octave Band Real Time Analyzer • FFT Plot (Fast Fourier Transform) • Linear...apps.apple.com
Max+Min, 2048FFT, linear
If you removed the logic board and all internals from the case and had it sitting in the open in a massive room, it would run considerably worse in reagrds to cooling capacity and noise. This isn't just likely to occur, it will occur, due to the design and you explain why next
This is what I was getting at a few pages back and is exactly right. The case size has not much to do with the thermal capacity, rather its design is what impacts how well it can cool. In your example the case directs the airflow across the heatsink fins, without the case it doesn't so in an open room it would run max fans and throttle badly.
You couldn't be more right. And due to a better designed cooler between the M4 and M4 pro, one being copper, a larger heatpipe and double the amount of fins (twice the surface area), then at exactly the same fan speed=airflow=cfm=noise, the cooler in the pro will remove much more heat because it is better designed to do so. Which in turn allows the processor to work harder without generating any extra noise.
No the main issue here is the effective cooling, so b), the case volume is not the issue. Like the example above where in an open room the system will perform much worse, if you moved the internals from the new small mini into the old larger mini, it would perform worse but not as bad as out in the open. Becasue it is the case and thermal design as a whole, not the case size that affects cooling performance. That and also the laws of thermal dynamics are kinda important.
I mostly agree with this too, a larger case does afford more room for a larger fan, though I'm confident you can have better cooling performance in the same small chasis with a better design, but as you said it likely wouldn't market as well as it isn't as slick with cutouts. There actually looks to be room for a larger capacity fan in that current design but they chose to use what appears the same size fan as all their machines (already had the parts I guess), that and a rework of the heat vent and it could've had much better cooling capacity. Though it would have to have a big hole in the side to vent heat out instead of from the foot. Apple did make an effort though, the two heatsinks definitely have different heat handling capacity, but compromise was made. If they kept the original chasis size, then yes it affords more leeway for cooling design with the extra space.
For the majority of people it will run fine and they will never even spin up a fan to hear it and there will be those that this is just too much for, for them they built the studio. I think a lot of people are confusing that it is a Mac Mini with the M4 pro chip, not a Mac mini pro. In saying that maybe they have confused the marketing message on this, maybe they should have stuck with Mac mini M4 and Mac studio M4 pro and max. But then people whine we had an M2 pro but no M4 pro in the mini.
I have a feeling there will be a firmware update to tweak the fan curves and profiles, maybe even vcore to the cpu to control temps, to better manage this as I'm sure the "noise" it generates amongst users will build. In any case I'm still getting one to try during extended returns.
The logic in this thread regarding the size of the device housing is not entirely comprehensible. Anyone who has ever assembled smaller (x86) computers knows that it is very difficult to impossible to assemble a really small mini-ITX PC with high-end components that is whisper quiet under load. At the latest when you put in an RTX 4090, the thing will take off during the next gaming session (unless you squeeze in a water cooling system, which then of course requires a larger case). However, if you choose a larger case which allows the assembly of e. g. bigger, slower fans and water cooling components, the noise levels can be whisper quiet even under load. However, to achive this you'll need more space, as the components required to quietly cool the system simply take up more room in the case. Of course, the Apple silicon is much more efficient than a x86 relic, but the chip will still produce some amount of heat under load.
The argument for more space is not about using the freed up air space for cooling by itself - of course that doesn't do much. Rather, with more space you can implement a larger cooling solution that is better in relation to the perceived noise level under load. For example, you could install more heat pipes or use larger fans to transport the air to cool the components. Small fans are usually louder than large fans when using the same airflow.
It seems that the M4 PRO is not an ideal match for the Mini case, as the waste heat produced under load cannot be dissipated in a good ratio of cooling capacity to the noise generated. The Mac Studio that implement the scenario described above show that larger cooling solutions allow for quieter operation under load, which require the use of larger housings.
The argument for more space is not about using the freed up air space for cooling by itself - of course that doesn't do much. Rather, with more space you can implement a larger cooling solution that is better in relation to the perceived noise level under load. For example, you could install more heat pipes or use larger fans to transport the air to cool the components. Small fans are usually louder than large fans when using the same airflow.
It seems that the M4 PRO is not an ideal match for the Mini case, as the waste heat produced under load cannot be dissipated in a good ratio of cooling capacity to the noise generated. The Mac Studio that implement the scenario described above show that larger cooling solutions allow for quieter operation under load, which require the use of larger housings.
Another review that mentions thermals and fan noise.
Conclusion - Sure you can hear the Mini Pro fans under heavy loads, but the fan noise is not a real concern (better than expected). When doing tasks in the real world (i.e. running AI models, running final cut pro), you're probably not going to get the fans to kick in at all.
Conclusion - Sure you can hear the Mini Pro fans under heavy loads, but the fan noise is not a real concern (better than expected). When doing tasks in the real world (i.e. running AI models, running final cut pro), you're probably not going to get the fans to kick in at all.
Why do you think those are "high temps"? Do you know what the expected/correct operating temperature range is for the M4?
Well put.
There seems to be mainly two groups in this thread.
1: Someone that just want a good (not great) desktop computer for basic to medium tasks, and/or are shilling hard for Apple - trying to rationalize basically anything, because Apple can't do wrong (a bit harsh maybe, but well deserved in some cases IMO
).2: Someone who needs a powerful desktop computer - but mostly for the CPU, and who also value a good user experience and silent operation. A good example would be working with Logic Pro, using large projects (and needing 48-64gb of RAM for large sample libraries).
If you're in group 2, it's probably wise to wait for the new Mac Studio next year.
I had my mind set on the Mac Mini M4 Pro, and it's disappointing to keep waiting another 6 months - but these machines are supposed to last a long time, and they are expensive.
Last edited:
Haven't ordered my Mac Mini M4 - likely M4 Pro - yet, but my "fan/noise baseline" will be if there is a change compared to my currently used MBP 14" M1 Pro 10c/16c/32GB/2TB for my real world tasks, mainly building/compiling Java-based (backend) software with running containers + typical office stuff, here and there a bit photo/video editing, rare gaming (Starcraft 2 via Rosetta).
Except the rare gaming use case, fans are OFF all the time with the M1 Pro. With SC2 via Rosetta, fans spin up pretty quickly. Noticable? No.
For my dev use case, I could make them spin up to around 2400rpm (same as above, not noticeable, if TGPro wouldn't show me, I wouldn't know) when I build/compile multiple times close to each other over a prolonged time, giving the CPU no room to breath for a few seconds, but that's a completely unrealistic use case for me.
Do fans kick in and being even noticeable during benchmarks? For sure. Is this real world? IMHO not for my use case, but I'm not e.g. a professional video editor, where exporting stuff could lead to the mentioned "heavy load" for a prolonged time with noticeable fan noise. Whatever "heavy load" really means at the end.
Looking towards to more dev focused tests on Youtube (e.g. from Alex), or I simply order one and return if the little box is too noisy in the living room, which I honestly doubt.
Except the rare gaming use case, fans are OFF all the time with the M1 Pro. With SC2 via Rosetta, fans spin up pretty quickly. Noticable? No.
For my dev use case, I could make them spin up to around 2400rpm (same as above, not noticeable, if TGPro wouldn't show me, I wouldn't know) when I build/compile multiple times close to each other over a prolonged time, giving the CPU no room to breath for a few seconds, but that's a completely unrealistic use case for me.
Do fans kick in and being even noticeable during benchmarks? For sure. Is this real world? IMHO not for my use case, but I'm not e.g. a professional video editor, where exporting stuff could lead to the mentioned "heavy load" for a prolonged time with noticeable fan noise. Whatever "heavy load" really means at the end.
Looking towards to more dev focused tests on Youtube (e.g. from Alex), or I simply order one and return if the little box is too noisy in the living room, which I honestly doubt.
If you're in group 2, it's an open question, but assuming it's going to be loud instead of testing it, lacking any relevant data, is just stupid.
Your insistence that your Logic Pro use puts you in that group is in any case very questionable. You said earlier that it's stressing your old Intel Mini; if so, it's highly unlikely that it would place any sort of burden on the M4, Pro or basic.
With a base M4 mini I have the game on low graphics high textures (habit from the old days?), 4v4 with AI, approximately 1920x1080 windowed mode. It dips to around 80-100fps during moderate fights, 140-160fps early on just macroing up. Power consumption bounces around 19W-27W tops. Those aren't levels high enough to push the fan to spin up past the base 1000RPM level (which is essentially silent).