MacBook 16-inch Fan Noise

LeoKost wrote:

So, as per your say, it's Chromes fault that even just by a single tab of Google Hangouts video call, the Radeon High Side drains 20W connected to an external monitor, the overall temperature of the laptop reaches a high level and the fans start spinning at 3k-5k RPM. Correct?

There is something about your particular configuration that causes a 20W load to cause your fans to spin up that high, likely the combination of the GPU power load and the extra CPU load caused by Chrome - though I've no idea if your experience is any different with Safari.

If you look at the results I posted, when connected to my Dell U2717D, my GPU was drawing 20W for about a half hour and no fan was running much faster than 2400 RPM.

The fans have been experienced by many, but your screen going black and the machine resetting is definitely not normal and you should have your machine examined by Apple.

You should contact Apple Support directly for both issues:

Contact - Official Apple Support

William Kucharski wrote:

The thing is, for me there really isn't all that much heat; my fans are only spinning at 2400 RPM or so at most even when the GPU is cranked up.

William, could you test your machine again using the steps I outlined in this post?

Your previous tests used Safari to watch a single 4K video while browsing a website. I don't know many people who purchase a MacBook Pro and use it for two windows from a single app. Most people I know also use Chrome for improved compatibility with websites over Safari or simply out of sheer habit.

My own testing shows that Safari is extremely well optimized for power consumption (compared with Chrome, at least) and manages to keep CPU consumption way, way down. QuickTime also appears to be highly optimized. I'm not asking for a test against third party "unoptimized software" - only Apple's own tech. This is intentional as it keeps any "well, that software is clearly unoptimized" conversation out of the mix.*

* And this mix is silly. I have Chrome, VSCode, Discord, and Slack open on my 2018 MBP simultaneously. As I'm writing this post, this is my CPU history:

Three notes:



1. I have ~65 Chrome [83.0.4103.116] windows open with an average of 3-4 tabs per window.
2. I have 19 active Spaces and 2 fullscreen application windows across three monitors (LG UK650 @ 2560x1440; Dell 2408WFP @ 1200x1920).
3. I'm running Docker, which is also the single largest CPU usage consumer at 15-20% [unless I trigger Exposé gestures, which itself causes the WindowServer process to jump to the top].

I mention this to simply point out that even "unoptimized" applications shouldn't task the CPU enough to cause the fans to blast when used "reasonably". I would argue that my usage is outside the realm of reasonably and yet my CPU usage is still pretty low overall.

All,

I've done some additional testing using quality StarTech USB-C to DP cables connected direcly to USB-C ports on a raised MBP 16 inch 2019, 32GB RAM, Radeon 8GB, 1TB SSD with an open lid and Turbo Boost Switcher Pro (TBSp) with disabled Turbo Boost:

SATECHI = Satechi Type-C Dual Multimedia Adapter Dual 4K HDMI, USB-C PD, Gigabit Ethernet, Micro/SD Card Readers, USB 3.0

https://www.amazon.co.uk/gp/product/B07RH8L9GV/

UGREEN USB Hub = UGREEN USB C Hub, 9 in 1 Type C Hub with 4K HDMI, VGA, SD/TF Card Reader, RJ45 Ethernet, USB Hub, 100W PD Port

https://www.amazon.co.uk/gp/product/B076HN81GS/

KENSINGTON USB Hub = Kensington UH7000C USB 3.0 7-Port Hub and Charger, and Sync Hub with Power Adapter

https://www.amazon.co.uk/gp/product/B00O9RPXWU/

MBP LEFT TOP USB-C port -> Dell U2515H DP port (1), 2560 x 1440 at 59.88 Hertz

MBP LEFT BOTTOM USB-C port -> Dell U2515H DP port (2), 2560 x 1440 at 59.88 Hertz

MBP RIGHT TOP USB-C port -> Dell U2515H DP port (3), 2560 x 1440 at 59.88 Hertz

MBP RIGHT BOTTOM USB-C port -> UGREEN USB Hub (connected USB 3.0: Apple keybaord + moouse + WD Passport 3TB)

Conclusion: No significant change in drawn power by "Radeon High Side" and it's still ~20 Watt in iStats Pro (iSP) when using DP ove HDMI cables.

At the time usually have about 60 tabs open in 6 windows in latest Mozilla FF, Mail, Microsoft Remote Desktop 10, Ms Excel O365 with one spreadsheet (not too heavy), Skype, Whatsapp - all latest versions. Recently upgrade OS X 10.15.5 to 10.15.6 still using HDMI cables setup but no difference in behaviour.

• The average temperatures are similar 55°C/131°F-70°C/150°F with average 62°C/143.6°F in iSP whereas TBSp is off about 8°C/46.4° showing lower average 55°C/131°F.

• The fans have been spinning between 2000rpm - 2600rpm and ocassionally would reach around 3000rpm.

Previously used for nearly a month:

MBP LEFT TOP USB-C port -> SATECHI HDMI 60Hz -> Dell U2515H HDMI 1 port (1), 2560 x 1440 at 60 Hertz

MBP LEFT BOTTOM USB-C port -> SATECHI HDMI 30Hz -> Dell U2515H HDMI 1 port (2), 2560 x 1440 at 60 Hertz

MBP RIGHT TOP USB-C port -> UGREEN USB-C to HDMI adapter -> Dell U2515H HDMI 1 port (3), 2560 x 1440 at 60 Hertz

MBP RIGHT BOTTOM USB-C port -> UGREEN USB-C to USB 3.0 adapter -> Kensington USB Hub (connected USB 3.0: Apple keybaord + moouse + WD Passport 3TB)

All quality Neet HDMI cables 1m long and even swapped them for same brand 0.5m long to rule out any bad cables etc. SATECHI was connected initially to the RIGHT side MBP ports then switched over to LEFT with no significant diffrence.

I've spoken to Apple Business support and was refused a replacement even though COVID-19 prevented me to test my new MBP for over 4 months hence outside statutory 14 days to return the product. I'm not happy about the whole situation at the moment and as soon as I disconnect monitors Radeon drops to 2 Watts or even less like 0.11 Watts!

Also, I've bought a Arctic 12cm x 12cm 0.08 sone super quiet USB powered fan that's been initially coolling down touch bar area and recently placed behind to help suck out hot air. I've got an open monitor raiser and there is nothing obstructing MBP's fans so produced heat can easily eascape.

DP connected monitors:

No monitors:

Added 3 youtube videos tabs in FF playing at the same time on each external screen and enabled back Turbo Mode (played for about 5-10 minutes):

Will continue my tests and report back.

I did a SMC reset yesterday and thought it was better today, but this didn't worked for me. I saw it in another thread about this issue but it's still the same for me now... Fan is about 3000 rpm when doing a lot on my Macbook but with an external screen connected using USB-C... I hope this really can have a solution soon!

https://support.apple.com/nl-be/HT201295

ahmedfromreservoir wrote:

Although that's neither here nor there, the issue is to do with the AMD GPU that is so poorly designed it has to put its memory at the maximum clock rate out of fear that displays may flicker. This isn't an issue with nVidia GPUs or older AMD MacBook GPUs and is the reason for the fan noise and overheating of the laptop.

Fear or experience? I suspect AMD has tested far more monitor combinations than you have.

Did you read my earlier post denoting how similar clock issues occurred with a past NVidia GPU?

https://discussions.apple.com/thread/250878229?answerId=253215243022#253215243022

Once again, comparisons with older MacBook Pro GPUs is immaterial here as they were not driving GDDR6 VRAM to its highest clock speed.

The net result of all this uncontrolled throttling, partially fueled by the GPU eating into the cooling capacity of the laptop, is worse real world performance than an i5 from 2017.

In your application; there are a wide variety of situations where performance on the MBP 16 is better, simply because older MacBook Pros couldn't do what the 16 can at all or at the speed it can do it.

I get that in your situation, performance may not be what you hope, that doesn't make it true for everyone.

ryunokokoro wrote:

William, do you not understand how continuously pointing out your own experience in this thread causes them to feel as though you're discounting their experience?

Not at all.

Science - which computer science is - requires reporting of all results, positive and negative.

Further, the fact that my configuration does not experience the issue does not belittle those experiencing the issue, but rather shows there is some further, non-obvious factor that explains is part of the issue.

This helps Apple and anyone else examining the situation to root cause the problem at hand.

Knowing just the cases where a problem does occur does not help, the situation, especially if say Apple only saw my results when trying to duplicate it in a lab.

You and everyone else here may be able to see the issue, but unless it can be reproduced in-house, it will end up marked "cannot reproduce" as there's nothing to go on to reproduce it. Anyone who has ever developed and debugged software has been there.

I have given complete data about my configuration, which has aided both people here and at Apple in examining which configurations have suffered from this issue and which have not.

People find this thread looking for help; for a solution. When you chime in with "Well, it doesn't happen for me. My GPU draws 20W and yet my fans stay at a cool 2400 RPM!" there is an implied "***** to be you!" You may not intend this, but it is how it comes off.

That's your read of the situation, not the read of anyone looking at it in an objective, scientific way. I want to see who has the issue, who doesn't and look for the differences, as therein lies understanding of the issue.

Please, please, please stop belittling people's experiences in this manner. It would be much more helpful if you framed it as "I'm sorry to hear that your machine is having these issues. I'm lucky and don't seem to have this problem. That said, I'm in this forum to help, so here is what I would suggest that you try..."

If others stopped posting that this happens with any machine connected to an external monitor, I might.

I literally went out of my way to purchase adapters, with my own money, to try and reproduce this issue, only to find I couldn't. I'm not going to purchase a monitor next, but even that bit of information is very, very helpful to know, as is when some were able to reduce power draw by reducing refresh rates below those normally supported.

A response suggesting that people's expectations are too high is unhelpful, especially when you simultaneously claim that your individual machine does not suffer from the same issue. Please stop lording your awesome experience over people and, instead, spend your words here to help them.

That's just it; nothing we say or do here can truly "help" them, only a future software update from Apple can resolve their issue, and to that point I have continually stated that if you are experiencing this issue, that you should contact Apple Support directly; posting here has zero effect on resource assignment within Apple Engineering.

You have recently started pointing people towards the source of the problem (GPU clock frequencies as a result of particular monitor configuration[s]). Hurray!

Which is something we have only had the data to do in recent weeks, through the efforts of those who looked more in depth at why I have been unable to experience the issue myself in my configuration.

What you have yet to do is ask people here about the details of their setup; to follow up with them to see if they can try adjusting their resolution to find a happy middle ground. Rather, you seem content to say "GPU problem. As designed. Nothing you can do." Whether you actually say "Nothing you can do" or not that is how it comes off to someone reading it while looking for answers.

People have freely shared their configurations; I don't need to ask as I can't actually do anything with the data, only Apple can.

I've also posted explanations from AMD as to why they do what they do, and when that has been called "nonsense" have posted references to experiences with NVidia GPUs where they suffered from massive power draws where AMD GPUs did not, once again, due to clocking of VRAM.

It takes more effort to follow up with people and work with them on adjusting their settings. But that is also how you actually help. The more people report the specifics of their situation and how certain configuration adjustments produce certain changes, the more information others will have to compare. It's a win-win.

Please help people win.

I am; I told them to contact Apple Support, which is the only way this issue will be resolved, if it can.

[Part 2!]

literally went out of my way to purchase adapters, with my own money, to try and reproduce this issue, only to find I couldn't. I'm not going to purchase a monitor next, but even that bit of information is very, very helpful to know, as is when some were able to reduce power draw by reducing refresh rates below those normally supported.

Oh? That's awesome. Have you reported the specific adapters you tested and the results? The only post I've seen where you get into details about your setup is this post in which you listed a single adapter and two laptops (confirmed here).

What others have you tested? What was the power draw for the Radeon High Side when you used them? (I assume fan speed was at-or-below 2,400Hz, of course.)

A response suggesting that people's expectations are too high is unhelpful, especially when you simultaneously claim that your individual machine does not suffer from the same issue. Please stop lording your awesome experience over people and, instead, spend your words here to help them.

That's just it; nothing we say or do here can truly "help" them, only a future software update from Apple can resolve their issue

Wait, what? I don't understand. If a user can adjust the refresh rate, resolution, adapter, etc. and have their fans turn down, would that not resolve the problem (loud fans) for them?

You have already suggested that people turn off Turbo Boost. How is that not worse than simply suggesting that they try adjusting their display settings/connections to see if it has an effect?

You have recently started pointing people towards the source of the problem (GPU clock frequencies as a result of particular monitor configuration[s]). Hurray!

Which is something we have only had the data to do in recent weeks, through the efforts of those who looked more in depth at why I have been unable to experience the issue myself in my configuration.

Heh. That was me. ;)

People have freely shared their configurations; I don't need to ask as I can't actually do anything with the data, only Apple can.

What? Yes you can. You are clearly an extremely experienced user. You can help users by explaining how to adjust the settings in question. You could see that a user has a USB-C to HDMI cable and suggest to them that other users have found better success using DisplayPort rather than HDMI and suggest that they give that a shot.

I'm not suggesting that you can take this information and hack the GPU driver to fix the issue or anything. I'm suggesting that the more people report their configurations, the more we should see what adjustments seem to help people. You can encourage that to help the community. Several users were reporting their SwitchResX settings with varying degrees of success. SwitchResX is an application that lets you fiddle with the low-level details of the protocols that the GPU and display use to talk. It's far more detailed than simply Resolution and Refresh Rate. The unfortunate thing here? Those users didn't report which monitors and cables they were connecting with. The data is incomplete. This is where some guidance from a community professional might be able to help identify "the differences" that lead to a resolution to the issue.

---

Side note, I'm checking out of this discussion. I do not have the laptop in question and I do not have the time to continue responding. I hope that I have been helpful.

The central point of my post is that if you are experiencing this issue, contact Apple Support, as Apple are the only ones who can fix this issue or can conclusively state that it cannot be or that the risks that come with any potential fix outweigh the benefits.

People have used various third party utilities to try to work around the issue, but I wouldn't recommend any of them, as updates to macOS X may cause them to stop working, and certainly anything that modifies the speeds fans run at could conceivably cause damage that then wouldn't be covered by warranty.

I have posted the complete technical details of my "testing" on more than one occasion, and while I only tested with a single USB-C to MDP adapter I did need to purchase cables and such and spent $60 to test a configuration I don't actually use regularly myself.

I can't be responsible for how anyone perceives technical data presented as just that, and I am providing data solely in response to others making statements that are much greater categorically than they should be (they "all" do it, the MBP is "defective," etc.) and posters who challenged me to do so.

Those statements are as constructive as saying "you always" or "you never" in an argument; it immediately causes the "accused" to hunt for a single example to disprove the accuser.

Regardless, people are free to ignore or discount my posts, and many do - but those actively trying to gain a more balanced idea of what the issues here may be have, I hope, been enlightened somewhat.

I am not an Apple employee, I post here and have been doing so for over a decade because I enjoy helping other Apple users work through their issues, and providing concrete data is the best way to do just that.

There is also way to turn down the built-in display's refresh rate, to one of several lower settings including 50 Hz.

Change the refresh rate on your 16-inch MacBook Pro or Apple Pro Display XDR - Apple Support

Whether that makes enough of a difference to reduce the display RAM speed would need to be tested by users under varying conditions.

Have you tried reducing the refresh rate of the Internal display? or

disabling the Internal display with closed clamshell mode? or

disabling the (mostly marketing hype) turbo Boost with Turbo Boost Switcher?

Change the refresh rate on your 16-inch MacBook Pro or Apple Pro Display XDR - Apple Support

Use your Mac notebook computer in closed-display mode with an external display - Apple Support

http://tbswitcher.rugarciap.com

Spoke with Apple support because I was having the same issue with my Macbook Pro 2020 that I never had with my 2015 version.

I'm connecting to a Dell U3011 with a USB-C to Display Port cable. Had terrible fan noise and overheating.

Apple support had me reset the SMC (System Management Controller) and it seems to be much better. You can follow the instructions here: https://support.apple.com/en-us/HT201295

If anything changes, I'll update this thread. But for now, the fix seems to be working.

Actually I couldn't hear the fans at 2400 RPM with my head on the unit.

Certainly I have heard the fans before, usually when performing an install or other compute-intensive task.

I have also stated before, and will continue to state, if something about the operation of your unit bothers you, contact Apple Support directly:

Contact - Official Apple Support

The 16-in MacBook Pro is the "Muscle-car" of notebook computers. It has previously unseen performance. But since Intel is about three years late in delivering the smaller and less power-hungry chips it should have been built from, it is power-hungry and may be loud, especially when used with larger external displays.

For using a separate "travel" computer:

One way some Users deal with sync-ing files for travel is to put those files on iCloud, and share them back by using the same Apple-ID on the travel computer.

How to free up storage space on your Mac - Apple Support

.