MacBook 16-inch Fan Noise

You have unrealistic expectations as to how interactions with Apple Engineering will go.

Basically it's usually:

1. Apple Engineering reaches out to ask for more data.
2. You provide the data.
3. You will be thanked and will hear nothing more until a fix is announced at some future date.

Apple will not keep you informed, tell you what they are thinking or otherwise reach out to provide updates except maybe to ask you to test a possible fix. That's simply Apple's culture; even other groups within Apple aren't always updated as to what other groups are working on except on a need-to-know basis.

What you call "going MIA" is the way interactions with Apple Engineering work.

Particularly if it's something they can reproduce in-house, there's no need on their part for further interaction.

If you want that type of hand holding, you'd best choose another hardware vendor or an ISV prepared to offer that level of support.

Also, once again, if a post was deleted, it was because it violated the Apple Support Communities Terms of Use, one of which is that speculation as to the cause of issues or future products or software updates is forbidden as speculation.

These forums exist solely for Apple users to receive help from other Apple users, not from Apple, and not to report issues or concerns. Interaction with Apple, while it does occur, is on the level of a celebrity sighting at a mall in Los Angeles; it happens but should never be expected.

You and others here are having issues, that is clear, but your official channel is through Apple Support, not here.

Posts that violate the Terms of Use will be and will continue to be removed by the moderators, that's literally why they are here.

For what it's worth, I understand when driving the MBP 16 the fans go nuts and thermal shutdowns often result; ask anyone that's ever used a MacBook with an external monitor and they will tell you the exact same thing will occur; photo editing on an external 19" monitor with a MacBook was a recipe for thermal shutdown of a properly operating unit in less than ninety minutes, repeatedly.

Just because you can do something doesn't mean it can be done on an extended basis without consequence, and that's why it's both important to test a particular configuration and why Apple has a fourteen day full return policy.

From what it sounds like, you would be much happier with a new Mac Pro, which is specifically designed to operate with those kinds of loads on an ongoing and extended basis.

What gives me the idea its not a reliable machine?

1. My first unit dropped dead in 30min, there are users that got replaced 4 machines until they finally just requested refund.
2. my 2nd unit heats up to 70 while reading newspaper, fans didnt even respond correctly with system settings, or lets say i never saw cpu temp spikes from 50-95 for a second, like 98% idling, not a single app opened, only connected monitor, fans didnt even respond.
3. Oh yeah and 25+ pages here on this thread plus some of the threads like those:

https://discussions.apple.com/thread/250826263

https://discussions.apple.com/thread/250938529

https://discussions.apple.com/thread/250905859

Almost every first batch of new models had a malfunction, my 2012 i think - first 2 gpu model, nvidia chip malfunction where it was later confirmed there was a bad batch of nvidia chips resulting in desolving because of - also a flaw - bad heat management. - So they knew from the start, Apple managed to swep thousands of reports of this problem for 3 years, until they were forced to launch a repare program, because of the lawsuits. I was so lucky mine broke down one day before the repair program was launched - but there was many people that apple charged them 950$ for that logic board replacement.

So dude, should this give me confidence for using this machine? Saddly i am stupid yes as i have streamlined workflow in mac os, that is the only reason left to keep using this machines. Changing at this point feels like a lot of hustle, but i ll get there soon*er.

[Edited by Moderator]

Could you quote the user? Because I am the user that has 2 TS3+ docks and they both work just fine. There is no loud or runaway fans when having two 4k screens connected.

https://discussions.apple.com/thread/250878229?answerId=252137336022#252137336022

gmn39 writes:

I use CalDigit most of the time. So, I have the following:

1. No difference, high wattage and it is hot
2. Works properly (tested on 2K, 4K and a FullHD monitor), corresponding to the iStat consumes approx. 5-6 watts
3. The same as 1.
4. The same as 2. (the combination of displays make no difference)

In answer to:

1. One monitor is connected via CalDigit TS3+
2. One monitor is connected via CalDigit TS3+ in clamshell mode
3. Two monitors is connected via CalDigit TS3+
4. Two monitors is connected via CalDigit TS3+ in clamshell mode

fmorga wrote:

1- Fan Noise and Heat issues at MacBook Pro 16 Inches. 

2- All Models are affected. 

3- It looks like all MBP 16 are defected. 

4- The issue appears connecting external device to the USB-C. In special external display.

5- Issue is affecting by all kind of display 4K, Full HD.

5- Several test shown it is a hardware issue. (Test running with windows software shows same results).(Latest software updates have not improved the situation). Other users could consider a software issue.

6- The best action is to report to Apple. https://www.apple.com/feedback/

7- Till now Apple has not reported any issue.

8- It is a worldwide issue and not an special Lot of computers. From Poland to USA, Germany and Italy.... worldwide is reporting the same failure at fan noise and overheating at just browsing.

9- Some test shown the issue is in relation with the GPU. 

10- Apple support can not inform if the issue is resolved. 

11- It is recommended to return the MBP within the 15 days period of guarantee. 

12- All the users that have contacted and followed the indications of Apple support, for example : restart in safety mode and so on.. have not observed any improvement.

13-Bunch of people is waiting for a solution or communication from Apple in order to buy a new MBP. (Three months gone).. 

14- Several people reported that computer work almost properly with no external connections. 

15- Fan loud Issue improve if you close the lid. 

16- Some people reporting better behaviour if you connect MBP to a dock.  Some people reporting that there is not improvement even with connection to a Dock. 

17- Some people reporting the fans goes crazy just connecting external hard drive. This was my case connecting hard drive through a hub to the USB- C. Please confirm if still the case. 

18- Media reports: 

19- If I missed something, just added to the summary. Thank you

You have two 5s on the list

I already showed that not all models are effected. There is some 16" MPBs that are having issues as people have shown, but there are a lot more that are not having the fan issues.

Why is that not on the list for balance?

Ignoring that these are isolated incidents will make it harder for Apple to trackdown. Why? because saying that ALL are having the problem means they can test ANY, and when the problem of runaway fans doesn't show up it makes it look like you're just crying wolf, when there is a real problem with some people's devices.

Apple uses my 16" to get info when there are complaints in forums like this. When there is a problem they reached out to me and have run tests to see what it looks like inside with all of the data. They comeback and ask me if I want to test a certain patch before it's released to beta and can say yes or no. Guess what? The problem I had initially when I got the 16" (would freeze in when put in clamshell mode when connected to a thunderbolt dock) has now been fixed with 10.15.3.

Only a few people have put anything out here with real information rather than just "All models are affected". Having more numbers on the list has less meaning as most are just blanket statements that should be in paragraph form.

This could be narrowed down to something like this:

Some 2019 16" Macbook Pros are having issues specifically heat related without the system being taxed by user engagement. Here is what is known:

1. When external monitors are plugged in directly to computer the temperature increases - some as high as 90C. For some users unplugging it returns the temperatures down, for others the temperature does not drop nor do the fan speeds and required a reboot. The power usage increases from .08 - 7 watts to 20 watts on the Radeon High Side (RHS).
2. Those having a dock (TS3+) have demonstrated a power increase from .08-7 watts but that appears to be the difference between using the Intel integrated graphics vs AMD GPU. The fans did not increase much if any. For some users it did not seem to quiet the fans.

Too many issues being brought up will make it like throwing darts at a wall without a target. If there are other issues such as connecting an external drive causing fans to increase, they should be in a separate thread because of all the variables so this one can get solved, whether through software/driver updates or replacement hardware.

Since this is the most popular topic on the issue.

I have the same problem too. MBP 16" i7, connected to 2.5k display via HDMI, fans spin at 100% even with little load.

My regular routine does not load CPU more than 100-150/800% according to activity monitor. I also use intel power gadget and TG Pro to monitor CPU status and all temperatures.

I can't tell if fan noise problem appeared right from the purchase, but it it seemed to be here forever.

The last straw that held me from writing here broke yesterday: not only fans spun full blast but I experienced heavy throttling with CPU frequency dropped to 1 Ghz.

I believe this had something to do with weather, it was pretty hot out there.

Here's what I've found.

• none of temperature sensors have shown anything above 70ºC. Low 60s on CPU and GPU, airflow below 50ºC, battery 37ºC. I think theese are pretty regular temperatures for MBP.
• I've discovered powermetrics utility and it has shown me this:

CPU Thermal level: 178

GPU Thermal level: 100

IO Thermal level: 50

Fan: 5602.85 rpm

CPU die temperature: 65.95 C

GPU die temperature: 63.00 C

CPU Plimit: 2.00

GPU Plimit (Int): 0.00 

Number of prochots: 0

It looks like the entire system sets fan speed based on CPU/GPU Thermal Level parameters. I've not found any explanation on how this values are actually calculated. But they used to get even higher values at about 200/150. Even though all temperature sensors have shown quite low values around 65ºC.

This entire situation took place at my regular very moderate load.

Further digging gave me three common advices:

• turn macbook off, disconnect power cable, turn on etc, described here: https://support.apple.com/en-gb/HT204463;
• close the lid and use macbook with only external display;
• reset SMC.

Non of them actually helped. After rebooting CPU thermal level stayed at 100 and did not want to get lower even without load.

I'm really confused now.

Adapter:

• USB-C to Mini DisplayPort Adapter, ITD ITANDA 4K Thunderbolt 3 to Mini DP Adapter Cable for Apple New MacBook 2017, ChromeBook Pixel Samsung S8(No Thunderbolt 2)

Cables:

• Dell U2717D: Cable Matters Mini DisplayPort to Mini DisplayPort Cable in White 6 Feet - 4K Resolution Ready
• Apple 30" Cinema Display: Apple Mini DisplayPort to Dual-Link DVI Adapter

PhotogWithMac wrote:

Edit: okay I looked it up. The “discrete” GPU is the powerful extra GPU and the “integrated” is the low power standard GPU. So you somehow activated the discrete GPU without external displays connected. How did you do that?

Right. The "discrete" GPU is, in my case, the Radeon Pro 560X. The "integrated" one is the Intel UHD Graphics 630.

Certain scenarios will force the system to "turn on" (start using) the discrete GPU:

1. Use certain applications that request it (OBS is one such app).
2. Attach an external monitor.
3. Disable "Automatic graphics switching" in the Energy Saver pane in the System Preferences.

When I mentioned in my writeup that I tested a Handbrake encode with the discrete GPU enabled, I had used option #3 above to force the system to use the Radeon GPU.

PhotogWithMac wrote:

i now understand that handbrake had less power available because the discrete GPU was consuming power as well.

That is the wrong conclusion to draw from this. "Handbrake" doesn't consume power, it consumes CPU cycles. What it likely means (corroborated only through anecdotal "because I watched the CPU Frequency graph during the process" evidence) is that the CPU itself was throttled simply because the Radeon GPU was active. It wasn't doing any processing while Handbrake ran [as I recall], but the CPU still ended up running slower.

I chalk that up to the GPU adding heat to the system and the CPU clocking itself down to compensate.

William Kucharski wrote:

Since Chrome isn't an Apple product, it's not Apple's fault if your machine heats up running Google's wildly inefficient code.

I think you may be missing the point. Google's code may indeed be "wildly inefficient", but that doesn't mean that it constantly uses 100% of every core all the time. I posted a screenshot of my 2018 MBP's CPU history to show that even more than "basic" usage of Chrome doesn't cause the CPU usage to run wild.

The request that you test with Chrome serves two purposes:

1. Use a more common test configuration. Many users who report issues use Chrome.
2. Use less-efficient software to simulate a better 'common case'. Chrome at idle isn't dramatically worse than Safari. I posted a screenshot of my CPU History to show precisely this point. That it is less efficient is useful for us here as it allows you to quickly see if the difference between "silent" and "blazing fans" is a small margin of CPU or GPU usage [or not].

That's absolutely hilarious: don't use Apple's software, it's meaningless, use this third party software known to be a resource hog and that makes desk side Mac Pros run hot and has for its entire life as a product and see if your MBP heats - that's your argument?

No. My argument is that people use MacBook Pros for professional, creative tasks like running Final Cut, Motion, Premiere Pro, After Effects, Photoshop, and other software. I'm not asking you to open Final Cut and make a few edits to a 4K video. I'm asking you to test something less CPU-impactful than that, but more CPU-impactful than what amounts to "basically the lightest possible test".

I thought people were claiming just connecting a monitor and letting the MBP 16 sit idle would cause my fans to "go crazy." Now I have to be running specific software that uses excessive amounts of CPU to help heat the machine past what the GPU does alone?

For some this is the case. People in another forum have also shown that changing their low-level display settings with a third party application, changing their resolution, or swapping cables have resulted in improvements to the Radeon's energy usage. The question I'm asking you to verify here is "how much CPU headroom does your laptop have with its configuration?"

You've shown that connecting your monitors to your 16" laptop doesn't trigger your fans to heat up, even with a relatively high "Radeon High Side" usage. The obvious follow-up question is, again, "well, how much headroom does your machine have in this configuration then?" If your machine can somehow handle the same sorts of tasks that it does with no external monitor connected without fans kicking on, then that is very helpful information to have.

If your laptop idles with no fans but starts to spin up after a few minutes of more than trivial CPU usage, then I'd suggest that your machine suffers the very same issue that everyone here is reporting. Your machine may simply have a "lucky CPU/GPU" in that the CPU and/or GPU placed in your machine happened to be particularly "good" coming out of the fabrication binning process.

Streaming full screen HD video isn't "trivial," it's one of the more CPU-intensive thing we ask computers to do, video calls/conferences are another, editing 4K video streams is yet another.

Apple is known to offload video decoding tasks to special H.264 hardware decoders in the device, making the entire process even more efficient. Third party software doesn't do this. YouTube videos max out at 1080p on Safari as YouTube only offers H.264 up to 1080p. This limitation doesn't exist in Chrome/Firefox/etc., allowing you to view videos at 1440p or 2160p (4K) or beyond, but they also don't use the hardware decoding (at least at those resolutions). Someone even made a Safari extension that works around this issue by opening the video in another browser.

When I asked you to run the QuickTime portion of the test I specifically chose Apple software as I expect it uses the special H.264 hardware for its encoding tasks. It flexes different services than the browser, however, so it pushes things "a bit more" and in a slightly different direction than your basic Safari test.

The excuses as to why my machine doesn't do this seem to be coming fast and furious now, as if it's some magical unicorn despite no one else being able to prove their MBP 16 does see this issue when driving the same monitor in the same way.

Millions of Dell U2717Ds and Apple 30" Cinema Display are out there, it can't be hard for someone to give it a shot.

I don't know many people who would be interesting in running out and paying several hundred dollars to run a quick test. That said, if someone in this forum is willing to run said test, then I'm sure everyone here would love to hear about the results. The Apple 30" Cinema Display is legacy and no longer generally available. The Dell U2717D is available here for $400 USD, in case someone wants to give said test a shot.

William Kucharski wrote:

Thats what I use and have never had my fans even become audible.

How is this helpful to the conversation?

In general, I'm not sure how productive many of your comments are here, William. Rather than ask people about their configurations, their specific hardware, etc. you keep bringing up your "test" several pages back wherein you played an HD video on Safari and did some light browsing. I pointed out that the software you used in your test would hardly touch the CPU and asked that you run a slightly more stressful test but you refused. I also pointed out that while CPUs may share a SKU it doesn't mean that they all handle thermals the same way. It is entirely possible to "luck out" on a CPU (or GPU, for that matter) purchase. If you've ever looked into hardcore overclocking in the PC world, then you'd be familiar with this quirk.

However, as you've mentioned, the specific protocol used to connect the monitor(s) to the laptop do appear to have an affect. Under the hood, it appears that resolution, scaling, and refresh rate are all a factor. Others on a different forum have pointed to known issues with AMD's GPUs being fickle with their memory clock rates - if you don't hit just the right combination of hardware/software settings, then your GPU will simply run at its max settings (and, therefore, temperature), rather than a step or two down to something more reasonable. Given this quirk, then there may actually be nothing that Apple can do in software to fix the issue - the issue may be with AMD's architecture for this generation/configuration of GPU (excluding the 5600; more on this below).

And this may be what you're running into with your "lucky machine" (or rather "lucky setup"). Some people find their luck by switching from HDMI to DisplayPort. Others try that with to no avail. Some people are able to use special software to modify the deep details of their monitor's connection and find a set of acceptable settings that keep the GPU power draw low, others do not. It seems to be very relative to the hardware, with a not-insignificant number of configurations leading to the highest power generation.

Then there's the 5600 GPU upgrade, which people report fixes the issue for them entirely. I have not seen a single person report that this far more powerful GPU configuration has actually been worse (if you've seen this, I would love a link to the report!). The GPU memory hardware is apples and oranges here, but it's safe to say that overall GPU strength does not necessarily cause a commensurate rise in heat. Perhaps, though, this is a red herring and the actual issue was something else that Apple (or AMD) addressed with the new configuration. We are not privileged enough to know.

Regardless, I think the best thing that you can do to help others on this forum is to continue to provide explicit details (read: model numbers; resolutions; etc.) about the specific hardware you've used in your "no fans ever" setup and ask them to post theirs. If you provide people with links to the specific monitors and the cables you've used to connect them and then encourage them to post their setups, then perhaps the community may be able to start identifying a range of setups that will allow them to keep their powerful machine and use it as expected.

(I'm not saying that you need to keep retyping all of this - you can simply link people to your holy grail "configuration post".)

These details seem to be important. There is an AMD community post wherein a Windows user with a triple monitor setup found that they had to change their refresh rate from 59Hz to 60Hz across all three monitors before the GPU's memory speed dropped from 1250MHz to 150MHz and their system was able to run cool. (Note that they increased the number of times the GPU had to send data to the monitors and the change in timings allowed the GPU to lower its speeds.) An AMD spokesperson responded that this was "expected behaviour" (sounds familiar!). (They also allowed that the effect described can vary dramatically between GPU, BIOS, etc.)

To that end please list the following (brand and model, if possible):

• Your external monitor(s)
• For each monitor:
  • Cable/dongle/hub/etc. used to connect it to the laptop
  • Its resolution
  • Its refresh rate

I think your input here would be very valuable to the community.

ahmedfromreservoir wrote:

Ok I've been trying to avoid your comments because you keep getting people banned when they point out you're wrong, but I'll take a nibble,

I'm flattered that you believe I have the power to "ban" anyone, but I have told you and others I have no more power than any other user here.

If anyone has a post edited or removed its for violation of the Terms of Use you, I and everyone else here agreed to when you created a login.

You're basically the only person here who hasn't experienced the fan issue in your MacBook - I've been through four laptops, Tim has been through many different configurations, people here and elsewhere have problems - do you see what I'm trying to get at here?

Yeah, there's something unique to the configurations I have available to me that don't cause that issue.

I've yet to read of anyone duplicating my config that had their fans run away.

Perhaps if I had one of your monitors, I'd feel differently; I don't, so I can't say I've experienced it.

Note I've never stated that you or anyone else here didn't experience it.

1) More powerful nVidia GPUs don't have these thermal problems

If so, it doesn't matter given they aren't used in Macs.

The powerdraw of the GPU is inconsistant with the performance it's driving, the issue has already been pointed at the memory clock speed which for some reason pins itself to its maximum when driving two monitors

For reasons AMD alluded to in the post I copied from another forum a few pages back; they need to drive the VRAM at maximum speed to avoid flicker. That's the way their GPU is designed and they do what they need to to extract maximum performance for power. 20w is a lot, but it's far from the peak it's capable of drawing.

Apple released a new MacBook with the more powerful 5600M which doesn't have this problem

Because it uses a completely new and different technology, HBM2, that wasn't even available to the public in an AMD laptop GPU at the time the MBP 16 was introduced.

Technology marches on; it's like saying Apple admits their old laptops were bad because they released models using a new family of Intel processors that didn't exist when the current model line was introduced.

^ if you want to still argue please read those above again and again until you understand why you're wrong, especially focus on #2 and ESPECIALLY focus on #3.

I answered your questions above. Completely new technology that wasn't available to the public in a laptop AMD GPU before.

If you are honestly interested, you can read more about it here:

• ExtremeTech: HBM2 vs. GDDR6: New Video Compares, Contrasts Memory Types
• AnandTech: AMD Announces Radeon Pro 5600M Navi GPU with HBM2 - Inside Apple's MacBook Pro 16"

Other GPUs, including the much more powerful 5600M and older MacBooks GPUs don't have screen flickering issues and don't run at boiling point to drive external monitors.

Once again, a completely new technology that wasn't available in an AMD laptop GPU until June 15, 2020 is somehow better than an older technology, who would have thought that would ever happen in the computing world?

To refresh your memory about AMD's comments:

Multiple displays with different resolutions, refresh rates, timings and or using different display adapters/connections requires more resources from the GPU, this can move the GPU up into the next memory clock state to compensate and avoid issues such as flickering or corruption.

If all displays are identical, using the same resolution, refresh rates, timings and using identical display adapters/connections then the GPU may be able to run two or more without moving up into the next clock state. It can vary from Bios to Bios and GPU to GPU, but the expected behaviour is increased clock speeds so this is not something we can change.

https://community.amd.com/thread/214891?commentID=2793350

Kind of cute you moved on from arguing people should buy a Mac Pro to people should buy an iPad.

A poster earlier said they moved to an iPad Pro, which means silence is more important than ultimate performance for them.

For some others here, given what they do, they would be happier with a Mac Pro, though that unit obviously has fans of its own, and the GPUs used in tower machines draw a lot more power than the GPUs in laptops do, so odds are those fans will occasionally spin up too, just like the fan in an iMac or my 2010 Mac Pro.

The two things that bothers me about the Pro model - Apple you need to listen to this - is lack of USB ports. First, Due to space being locked in at purchase I edit video on an external SSD. If I'm plugged into power, and mostly likely you are if you are editing video, you use one port for that and one port for the SSD. Where do you plug in your dongle for reading the camera card? You would have to be constantly switching.

The second concern is lack of displays.

https://www.apple.com/macbook-pro-13/specs/

The video section says we are limited to one external display:

• "One external display with up to 6K resolution at 60Hz"

That is a problem for many pro users. You may as well stick with the Air and save the money if you are limited to one.

One day old MBP 16" i7 version.

My fans are spinning all the time and the top part of the machine gets warm to the touch.

I have noticed that Chrome is a problem since it by default will use the GPU and that has keep my previous two MBPs busy and hot. Settings -> System -> Use hardware acceleration when available is set to OFF and work pretty fine for my daily use (I'm a web-developer in JavaScript)

Never had any issues with Safari, but the developer tools stinks so my daily driver is Chrome.

My previous MBP 15" i7 mid-2018 was a bit smaller and didn't start the fans this often.

I just found that when the computer has a second screen, the fan will increase to 3000+ rpm, which makes it very noisey. otherwise it is very quite. I will change the HDMI cable to see if it is the cable's problem.

I'll join the chorus as one experiencing frequent and load fan noise. I came from a 2018 15" MBP that rarely made audible fan noise, and when it did, it was understandable (Premiere export, live streaming, etc.).

Granted, the new 16" MBP's loudest noise is likely not as loud as the 15", it is certainly occurring more often. I just opened Premiere—opened the app!—and the fans went on a 5 minute tirade. Most of the other occurrences are seemingly random and unexplainable as well.

There are other issues with this MBP as well that are off topic for this thread, but that are making me seriously consider returning the machine. They all seem comical for a ~$4,800 setup.

Hello.

I must also join the chorus about the MacBook Pro 16 inch's fan noise. Mine was received 10 days ago and I've connected it to two 1440p monitors. When doing anything with the machine the fans speed up. This is the somewhat same behavior I have experienced with my old late 2013 MacBook Pro 15 (only its other problems are more severe, reasons why I was looking for a replacement.) I also have a Mac mini 2018 which I planned on selling after purchasing the MacBook. When doing the same tasks on the mini the fan speed didn't go up at all. I am curious as to why the Mac mini with its much less body and worse GPU doesn't have a problem with connecting to two 1440p monitors but the MacBook Pro 16 inch does.

This inconsistency makes this machine seem like it is not reliable enough for me to invest over 3200 dollars (this is what it costs in my country) for the base model. This is why I am returning the MacBook.