Try charging on the right and not on the left(apple.stackexchange.com) |
Try charging on the right and not on the left(apple.stackexchange.com) |
Did anyone ever figure out why exactly the temperature sensor was making kernel_task go crazy?
There was an app called Rain that you could download and it would just keep the idle process from running away and causing heat. I may need to research this on my lunch break and report back.
EDIT: Here's a related thread: https://arstechnica.com/civis/viewtopic.php?f=8&t=1060257
Should I understand that having any kind of load on the measly 2 ports I have on my 1600€ computer will lead to abysmal performance? Like, attaching a USB hub to add basic necessity like HDMI or USB A?
As I understand it, though, most of the thermal issues are to do with the north bridge and the video card. Basically, Intel's power model for USB is built around performance, not thermals. GamersNexus actually has some interesting videos about this, but at the end of the day Intel shit the bed on northbridge performance while developing USB-4 support.
I wildly speculate that this is a massive part of why Apple moved to their own silicon: to get away from Intel, who is clearly losing every competition they're in. (I don't know why Apple didn't go toward AMD, but I expect GPU hang-ups were part of it.)
- Leave the laptop screen open. You can turn off the screen by setting its brightness to the minimum setting.
- Power your USB devices from a powered USB hub.
- Point a small fan at the side of the computer.
- Clean the laptop's cooling system of dust and lint. Keep the laptop on a stack of books so the intake vent is raised above any surface that can collect dust.
Is there any actual evidence for running a CPU near its thermal ceiling causing premature failure vs a cpu running at a 10-20c lower temp?
When I hear people saying things like “my PC is dying”, this is usually due to thermal throttling from dust buildup or malware running causing excessive CPU usage. It’s not like the CPU is getting crispy around the edges and parts of it stop working so only a smaller portion is still available to do work or something.
There seems to be a lot of conjecture out there about the longevity of a CPU vs heat, but I’m wondering if this has ever been actually studied in a scientific way. I understand that electromigration is real, and it could e.g. cause a trace to eventually blow out.
This is the closest thing to a real scientific explanation that I could find https://www.reddit.com/r/hardware/comments/5l3ufj/is_there_a... but it still doesn’t go into why / how / the physical mechanics of it.
https://forums.tomshardware.com/threads/cpu-electromigration... Also attempts to explore this in a fairly scientific manner, but it’s focused on over-volting and overclocking vs normal (especially mobile, where CPU voltages tend to be less tweakable) operation.
TL;DR In my experience, CPU/GPU temps don’t matter as long as they are not causing throttling. Running within a few degrees of the thermal limit (Tjunction max ) 24x7 for years won’t affect the longevity of a CPU/GPU in any functional way, and won’t cause slowdowns, only failure, possibly of a subsystem like USB or other on-die subsystems. If anyone has anything showing otherwise, please share it!
Benchmarks show the MB having decent performance, as they start cold and don’t have external devices attached either. Thermal throttling only shows up once you’ve already bought
Running CMake + gcc on a bigger project? The fans start running like crazy. The performance of the i9 is awesome, at least for the first 1-2 minutes, after that the CPU is throttling you back to the 90's and the fans are spinning even louder.
But this is not even the funniest issue with it. I have the biggest and most performant charger (96W, pretty much the maximum, because the USB-C standard states, you can only push up to 100W through it). You may think, if Apple put that charger in the box, it should be enough to achieve max performance with it AND still be able to charge it. Nope, but maybe it would be at least enough power, so the battery is not discharged. Lol, nope, higher load for a extended time WILL discharge your battery while being connected to the power supply.
I've tested it with two separate MBP with exactly the same spec (i9 + 32GB RAM + Radeon 5500M 8GB) with exactly the same behavior.
No. Apple sells computers regardless, but now that "form over function" Johnny Ive has left Apple, Mac sales have actually shot up. People want their computer to look and feel good, but it still needs to function as a computer. Now that the actual functionality of the Mac is a priority again, they are selling better.
Maybe they fixed this issue in the time between 2016 and 2019.
My late-2020 MacBook Pro also stays warm (even hot, sometimes) while supposedly sleeping. It can also no longer charge on the right side, that just stopped working. Had to rearrange my desk to satisfy the power bugs on the laptop.
Quality has gone way down, I don't quite see myself ever getting a Mac laptop again, after being on OSX laptops ever since OSX 10.0 came out.
I mean, it's happening cause charging it, that's something all users will do :S
Sounds a lot like the antennagate
Also why is that submission now marked as a dupe? It doesn't even point to the one that it is supposed to be a dupe of.
And the noise is not much worse, because eventually they spin at full blast anyway; just a bit too late when presumably thermal throttling is already hitting.
Everytime I connect an external monitor, it makes the GPU go crazy and overheats the Macbook enough to summon this kernel_task demon.
Couldn't comprehend why one of the best laptops on the planet cannot handle one external monitor.
The Pentium IIIs were the first CPUs that could consume enough power to cause heat problems if not throttled. OSes didn't yet anticipate that; they didn't yet have the concept of issuing halt instructions when there was no application work to do, they just let the CPU spin faster on a waiting-for-message loop. The OS makers soon recognized the need, and halting the CPU became built in to the OS by the time of either Windows 98 or 2000, so yeah, in the interim, tools like Rain were what you used.
On processors that have a halt instruction that stops the CPU until an interrupt occurs, such as x86's HLT instruction, it may save significant amounts of power and heat if the idle task consists of a loop which repeatedly executes HLT instructions.
--
I don't know how it is implemented in darwin, HLT is a privileged instruction, then again it's kernel_task.
https://tidbits.com/2020/06/17/how-to-report-bugs-to-apple-s...
I'll go further and I'll say I think it's actual dishonesty on their side: they're providing additional thermal headspace to their own cpus, when instead they have constrained Intel for years.
It's still pointless to point this out, apple users are going to buy apple stuff anyway...
Edit: why is this downvoted? This is a legitimately confusing mistake; people may think it has something to do with synchronization.
You ought to have commented: "I believe you meant heat sink, not 'heat sync', am I correct?"
The average HNer is above average in intelligence and does not require a minor typo to be explicitly corrected.
If it's about CPU temperature, why does Apple schedule a task instead of just downclocking the CPU, like everyone else? This question predates the release of M1 Macbooks, and everyone else does thermal management of Intel Core processors via clock speed, not artificial idle load.
It's annoying that the process monitor does not provide any way to distinguish high CPU load due to demand and high CPU load due to a low clock frequency but there is software that can display a CPU frequency chart which can give some insight:
https://pietschsoft.com/post/2020/02/15/macos-monitor-cpu-us...
If you're on Linux, then the "tiptop" utility does just this.
There's similar behaviour on linux but it shows as specific kernel threads related to intel power throttling, so it's more obvious what's happening.
The M1 can't even support two monitors plugged in directly, which is a joke for a machine this expensive.
That is the case for the M1, which is the lower-end of the new processors – a bit disappointing because I'd've otherwise bought one, but a relatively uncommon scenario overall for a lower-end Mac.
It's not the case for the recently-launched chips, which support up to 2x6k external displays (M1 Pro) or 3x6k + 1x4k (M1 Max).
OT: It’s nice that DisplayLink connections let an M1 machine have unlimited displays, and that the performance of their driver under Big Sur and above has been fantastic for me for the past half year for 2 extra displays. (Four total, 1 is vertical)
Implying Apple's cooling is any better? Both build difficult to repair devices tuned only for looks. Whether it's the Macbook that literally doubles its performance if you build a custom cooler that prevents throttling, or the Surface Book that you can't use at full performance for long because the charger cannot deliver enough power to sustain it.
Nobody can cheat thermodynamics, and attaching "Pro" to the name doesn't make it go faster either. I just hope I can keep buying "normal" laptops. They're ugly and noisy, but at least I know they can be fixed without melting off the keyboard or the repair guy telling me it's water damage when it isn't.
Intel integrated GPUs are often set up to use main memory instead of dedicated GPU memory (as for the Iris graphics). I wonder if that makes a difference, though it is on the same package.