MacBook Pro Fan Thermals not as “advertised”

[raftman]

I just want to share my MBP experience with thermals and fan noise. The YouTube reviews I watched said the fan is silent, it’s hardly even spinning even under max CPU load. My experience is MUCH different.
I’m playing Fortnite in “windowed mode”. And after 30 minutes, the fan is easily audible and you can feel lots of hot air blowing out the back. For some reason “windowed mode” causes this more.

My feeling is the Youtubers aren’t maxing out the CPU and GPU simultaneously for that long. My fan RPM is much higher with pretty average gaming.

Anyway my final thoughts on the cooling are that the heat pipe and fan work very well, and are sized correctly for this CPU when you actually push it to its limit.

 

[bill-p]

Yeah, max CPU load is about 7 - 10W, and max GPU is about 7W. Max for both about 15W.

7W should not need a fan, but 10 - 15W should. That's normal.

I suspect this is less because Youtubers are not stressing the machine enough, but rather because we're in winter. Room temp is now low enough that most folks won't really need the fan at all.

Wait until summer and we'll see. I suspect the MacBook Air will throttle quite badly, while the Pro won't throttle but its fan will be quite loud.

 

[dpgx81]

No issues here so far, I haven't taxed it but its nowhere near as hot/loud as my 16" gets.

 

[raftman]

Yeah, max CPU load is about 7 - 10W, and max GPU is about 7W. Max for both about 15W.

7W should not need a fan, but 10 - 15W should. That's normal.

I suspect this is less because Youtubers are not stressing the machine enough, but rather because we're in winter. Room temp is now low enough that most folks won't really need the fan at all.

Wait until summer and we'll see. I suspect the MacBook Air will throttle quite badly, while the Pro won't throttle but its fan will be quite loud.

I also suspect that the Air throttles more than the reviewers say. I mean there tests don't even heat up my Pro, compared to my real world use. Also judging by how much hot air is blowing out the back.

 

[Nebrie]

The reviews I watched said that the fans don't turn on for the first 10 minutes. I've never watched anyone claim that the fan stays off under load for 30 minutes. Compare that with an Intel Mac where the fans immediately ramp up to jet engine levels the second you tap the CPU.

 

[raftman]

The reviews I watched said that the fans don't turn on for the first 10 minutes. I've never watched anyone claim that the fan stays off under load for 30 minutes. Compare that with an Intel Mac where the fans immediately ramp up to jet engine levels the second you tap the CPU.

I’m talking about fan noise. I watched many reviews that said you can’t even hear the fan. Where as I can easily hear the fan from 10 feet away. The reviewers never got the fan RPM up compared to my usage.

 

[nameste]

I’m talking about fan noise. I watched many reviews that said you can’t even hear the fan. Where as I can easily hear the fan from 10 feet away. The reviewers never got the fan RPM up compared to my usage.

I think there are some honest YouTubers

 

[dingclancy23]

I just want to share my MBP experience with thermals and fan noise. The YouTube reviews I watched said the fan is silent, it’s hardly even spinning even under max CPU load. My experience is MUCH different.
I’m playing Fortnite in “windowed mode”. And after 30 minutes, the fan is easily audible and you can feel lots of hot air blowing out the back. For some reason “windowed mode” causes this more.

My feeling is the Youtubers aren’t maxing out the CPU and GPU simultaneously for that long. My fan RPM is much higher with pretty average gaming.

Anyway my final thoughts on the cooling are that the heat pipe and fan work very well, and are sized correctly for this CPU when you actually push it to its limit.

30 minutes on Fortnite is really good. When I launch Fortnite, The Intel fans kick in within 20 seconds.

As people get used to this type of performance, what was once amazing two weeks ago will now feel average. But then they will not be able to go back to their previous laptops.

This will be similar to when Retina was first introduced. Once you have seen it, Retina is normal and you cannot go back.

 

[Frantisekj]

Faulty unit?

 

[jdb8167]

I also suspect that the Air throttles more than the reviewers say. I mean there tests don't even heat up my Pro, compared to my real world use. Also judging by how much hot air is blowing out the back.

My M1 MacBook Air is faster than my 6 core 12 thread 2013 Mac Pro at doing video compression using Handbrake and ffmpeg x265 for the compression. If it throttles it isn’t noticeable and apparently doesn’t slow it down much. I’ve run the video compression on high quality for movies that take over an hour to finish. The Air barely gets warm.

 

[raftman]

I think there are some honest YouTubers

Yes, that YouTuber is exactly right, good find. I think the reviewers copy each other. So somebody said it’s completely silent even under heavy load, then everyone stuck with that.

I’m very happy with the level of fan noise Though. 100x better than the 2020 MBA i5 that I bought and returned. The fan on it would blow COLD air out the back at max RPM, what a crap design that was.

 

[Tubamajuba]

The reviews I watched said that the fans don't turn on for the first 10 minutes. I've never watched anyone claim that the fan stays off under load for 30 minutes. Compare that with an Intel Mac where the fans immediately ramp up to jet engine levels the second you tap the CPU.

And when there's no fan, like in the newest Intel MacBook Air, the damn thing idles at like 50-60 degrees celsius and gets into the 90s just by opening a web browser.

My M1 MacBook Air is faster than my 6 core 12 thread 2013 Mac Pro at doing video compression using Handbrake and ffmpeg x265 for the compression. If it throttles it isn’t noticeable and apparently doesn’t slow it down much. I’ve run the video compression on high quality for movies that take over an hour to finish. The Air barely gets warm.

That's insane. Insanely awesome.

 

[nameste]

Yes, that YouTuber is exactly right, good find. I think the reviewers copy each other. So somebody said it’s completely silent even under heavy load, then everyone stuck with that.

I’m very happy with the level of fan noise Though. 100x better than the 2020 MBA i5 that I bought and returned. The fan on it would blow COLD air out the back at max RPM, what a crap design that was.

Well i still can’t decide between mba or mbp.I was really uncomfortable with my i9 16” 5600m fan noise so I decided to return it. I think I can handle some noise but if it starts to make the noise you mentioned I think I can’t handle that.Is there a way you can test it with an dB app?

 

[deeddawg]

I suspect this is less because Youtubers are not stressing the machine enough, but rather because we're in winter. Room temp is now low enough that most folks won't really need the fan at all.

Wait until summer and we'll see. I suspect the MacBook Air will throttle quite badly, while the Pro won't throttle but its fan will be quite loud.

What kind of indoor temperature swings are you seeing where you live?

In these parts indoor temperatures are typically ~72F winter, ~78F summer. I'd be surprised if 6F ambient made any appreciable difference.

... and it is almost summer in Australia and other southern-hemisphere nations. I don't follow youtubers but presume at least one or two likely live outside the US?

 

[nameste]

Here is a YouTuber lives in Australia

 

[bill-p]

What kind of indoor temperature swings are you seeing where you live?

In these parts indoor temperatures are typically ~72F winter, ~78F summer. I'd be surprised if 6F ambient made any appreciable difference.

... and it is almost summer in Australia and other southern-hemisphere nations. I don't follow youtubers but presume at least one or two likely live outside the US?

San Jose, CA.

This past summer saw 110F. Right now, it's barely 45F in my room. My 16" MacBook was screaming for dear life until I turned on the AC.

Here is a YouTuber lives in Australia

Was going to say... that's not an intensive workload at all. The new M1 chips have separate video encoder/decoder processors that barely cause the CPU to go up at all during export/playback. Those processors likely won't cause much heat in the system.

Try running both Cinebench R23 and GFXBench outside like that. Or run Fortnite @ 120fps. I think we'll see a different story.

Also, maybe it's just me, but this is like... 10th video I've watched where a youtuber claims amazing FCPX performance and also only does thermal testing with it. I mean... I get it. They're youtubers. 101% of their workload is FCPX. But it's no longer the most intensive workload with M1 chips, as opposed to how Intel MacBooks are handling it (mostly dependent on the encoder/decoder processors on the GPU).

 

[deeddawg]

San Jose, CA.

This past summer saw 110F. Right now, it's barely 45F in my room. My 16" MacBook was screaming for dear life until I turned on the AC.

Here in my part of Georgia (US) we infrequently hit 100F, but we do spend much of the summer exceeding 90F. Fifty days this year. It is a rare building that doesn't have AC.

My dear wife would beat me senseless if I let the indoor temp go down to 45F while we were present. 72F is our compromise; I'd be happy with 70F in winter, she'd rather have the thermostat at 74F.

 

[raknor]

San Jose, CA.

This past summer saw 110F. Right now, it's barely 45F in my room. My 16" MacBook was screaming for dear life until I turned on the AC.

Do you live in a tent? There is no way internal temperatures hit 110F or 45F even in the worse insulated houses in San Jose. I used to live in a house built in the 1940s single pane windows and never saw those indoor temperatures in the Bay Area.

 

[dingclancy23]

San Jose, CA.

This past summer saw 110F. Right now, it's barely 45F in my room. My 16" MacBook was screaming for dear life until I turned on the AC.



Was going to say... that's not an intensive workload at all. The new M1 chips have separate video encoder/decoder processors that barely cause the CPU to go up at all during export/playback. Those processors likely won't cause much heat in the system.

Try running both Cinebench R23 and GFXBench outside like that. Or run Fortnite @ 120fps. I think we'll see a different story.

Also, maybe it's just me, but this is like... 10th video I've watched where a youtuber claims amazing FCPX performance and also only does thermal testing with it. I mean... I get it. They're youtubers. 101% of their workload is FCPX. But it's no longer the most intensive workload with M1 chips, as opposed to how Intel MacBooks are handling it (mostly dependent on the encoder/decoder processors on the GPU).

I do not see the point. The previous Macbook Airs cannot run Fortnite at 30fps much less than 120fps. The goalposts and expectations for these devices are moving.

FCPX is also another case. It is not the M1's fault that it has hardware acceleration and now suddenly it is not a valid test because the Intels do not have it. The device is the sum of all its parts.

Just remember the baseline we are working with here (AKA Intel).

 

[RigSatMe]

My M1 MacBook Air is faster than my 6 core 12 thread 2013 Mac Pro at doing video compression using Handbrake and ffmpeg x265 for the compression. If it throttles it isn’t noticeable and apparently doesn’t slow it down much. I’ve run the video compression on high quality for movies that take over an hour to finish. The Air barely gets warm.

It is obvious: M1 has 8 cores with hardware codec support for HEVC, but Mac Pro 6 cores, which doesn’t support HEVC on hardware level. The same gonna happen with M1 when h.266 is released.

 

[Zandros]

It is obvious: M1 has 8 cores with hardware codec support for HEVC, but Mac Pro 6 cores, which doesn’t support HEVC on hardware level. The same gonna happen with M1 when h.266 is released.

x265 is a software encoder, it should not be using any fixed-function hardware in the SoC. The decoding might be accelerated, which helps a lot in some cases, but we have not been told what the input is.

 

[jdb8167]

It is obvious: M1 has 8 cores with hardware codec support for HEVC, but Mac Pro 6 cores, which doesn’t support HEVC on hardware level. The same gonna happen with M1 when h.266 is released.

I wasn’t using videotoolbox. I was using the videolan x265 codec. No hardware codec support. It does use NEON acceleration on the CPU.

 

[jdb8167]

x265 is a software encoder, it should not be using any fixed-function hardware in the SoC. The decoding might be accelerated, which helps a lot in some cases, but we have not been told what the input is.

The input was raw mpeg ripped from a DVD. A .ts file.

 

[bill-p]

Do you live in a tent? There is no way internal temperatures hit 110F or 45F even in the worse insulated houses in San Jose. I used to live in a house built in the 1940s single pane windows and never saw those indoor temperatures in the Bay Area.

No, just a regular house. That last heat wave that hit the Bay Area a few months ago caused record high temperatures. You really have to live here to feel it.

Bay Area faces another potentially record-breaking heat wave with temps soaring into 100s

Here we go again, Bay Area! Scorching heat wave heading our way will bring triple-digit temperatures and dangerous fire conditions, says Meteorologist Drew Tuma.

abc7news.com

It's not "usual occurrence" per se, but it did teach me that I'll always need a fan in my computer, just in case.

I do not see the point. The previous Macbook Airs cannot run Fortnite at 30fps much less than 120fps. The goalposts and expectations for these devices are moving.

FCPX is also another case. It is not the M1's fault that it has hardware acceleration and now suddenly it is not a valid test because the Intels do not have it. The device is the sum of all its parts.

Just remember the baseline we are working with here (AKA Intel).

Again, the point is simple: if I'm paying for a device, I want to make sure I can use it at its best, all the time, anywhere, regardless of what kind of workload I throw at it.

The test is too "selective" because it shows me I can export video under direct sunlight for 5 minutes. But it doesn't show me if I can stitch a 1GPixel panorama photo, or play an intensive video game under the same condition.

In fact, why dwelve on the old Intel Macs now? Pretty soon, these M1 Macs will be the baseline. So what if the M2 Air in the future thermal-throttles and performs worse than the M1 Pro when ambient temperature is too high? Will you then say "but Intel Macs" again?

 

[thelookingglass]

This past summer saw 110F. Right now, it's barely 45F in my room. My 16" MacBook was screaming for dear life until I turned on the AC.

Not that I'm doubting you or anything, but if your room is 45F right now you've probably got on a sweater and some sort of winter coat on top. Are you sure it's 45F? lol