Crazy heat on MBP '15

[Thysanoptera]

The power flat lines too though. And it flat-lines exactly at or around the chips TDP, which makes perfect sense. The chip simply stops boosting further and an equilibrium is reached. I don't see any evidence for sensor saturation, but maybe I didn't completely understand what you mean.

I'm talking about the initial spike lasting seconds, to 80-90W, where temp stays magically flat at 100C and starts dropping only after the power crosses back 50-40W on its way down. Only after that it bounces back and stabilizes at 100C around 45W.

Oh you could but you had to find the way around Intel's clock limiter Overclockers have been doing this stuff on unlocked CPUs for ages. And Coffee Lake is almost an unlocked CPU since Intel put the limits much higher then on previous generations...

Yes, I know, but we're talking here about Apple laptop and not purpose build overclocking rig. If you know about a way to overclock Apple product, or even getting 2 x TDP out of Clevo barebone with Prema bios, I'd be more than happy to hear about it.

 

[MacBH928]

My current MacBook Pro on the other hand has zero heat issues, and since this thread discusses the CPU, having run my CPU at near 100C for many, many, many hours on all of my Macs; 0 issues so far. Going far back mind you. I don't know when I first started keeping an eye on these things, but I've been doing similar tasks to what I do now on my computers basically since the G5 - Maybe little bit on the G4 iMac as well, though that was probably a hell of a lot lighter.

which MBP you have? How long have been using it? I will be amazed to know that at 100C for hours at a time won't affect the internals
[doublepost=1550499043][/doublepost]

No, it is not normal and fine. No. No no no. 100C is the maximum temperature the intel chips are rated for, and if you are hitting 100C (I have never seen this, not in 15 years) it means that you are seeing hardware-induced throttling happening so the chip doesn't fry itself. Again, no. I have no idea where you got this stuff.



leman is clearly a smart guy, which makes it more depressing when you see someone intelligent being willfully deceptive.

If you own a mbp try going over to mixer.com and watch a streaming video for five minutes... tell me what temps you are reaching.
note I can tell the temps of the CPU because I use software called iStat Menu... i know of no other way to monitor the CPU heat

 

[SDColorado]

ALL 15" MBPs produced in last 6+ years will reach 100C when doing intense work, because that is how they are designed to operate. It's normal and there eis nothing wrong with it. The machine is built to keep the CPU operating at 100C when it draws 45W of power. It's that simple. And if you current MBP never saw this temps it simply means you never did anything remotely demanding with it.

I have hit 97/98c frequently, but not 100c before the fans really kick in and it will hover around 96/97c when pushed for an extended period. But the previous one I did return, used to hit 100c frequently doing even doing random nonsense tasks like syncing my phone via iTunes or opening mail. I would conclude this one is working correctly.

If you own a mbp try going over to mixer.com and watch a streaming video for five minutes... tell me what temps you are reaching.
note I can tell the temps of the CPU because I use software called iStat Menu... i know of no other way to monitor the CPU heat

Do you have a specific streaming video in mind on mixer.com? I watched a couple last night and a couple this AM and am only seeing temps in the low 60c's. Perhaps I am watching the wrong thing?

 

[Eason85]

which MBP you have? How long have been using it? I will be amazed to know that at 100C for hours at a time won't affect the internals
[doublepost=1550499043][/doublepost]

If you own a mbp try going over to mixer.com and watch a streaming video for five minutes... tell me what temps you are reaching.
note I can tell the temps of the CPU because I use software called iStat Menu... i know of no other way to monitor the CPU heat

I don't own an MBP anymore, but I went there with my X1 Extreme for funsies and saw 62C. That's a similar CPU (i7-8850H) in a 3.7 lb laptop.

 

[maflynn]

ut I went there with my X1 Extreme for funsies and saw 62C

I have the 8850/4k X1 Extreme. My temps are in the low 40s doing normal work, surfing. Streaming they perk up to the 50s and playing Fallout 4 (the only game I'm playing on it) push the temps to the upper 70s I've yet to see it it 100c.

 

[chabig]

I will be amazed to know that at 100C for hours at a time won't affect the internals

People are amazed by many things. That doesn't make it true. If the internals were at risk by 100C heat, then Apple engineers would have built the machine to keep the temperature lower.

 

[maflynn]

f the internals were at risk by 100C heat, then Apple engineers would have built the machine to keep the temperature lower.

That's a huge assumption that imo may be more of a leap of faith. It doesn't matter who made the laptop, I do not want to see 100c consistently for that very fact.

 

[chabig]

That's a huge assumption that imo may be more of a leap of faith

This thread isn't really going anywhere because everyone is just posting opinions. The only fact under discussion is that Intel's published maximum CPU temp is 100C. If people are going to believe that Intel is somehow lying in their specifications, then we are delving into conspiracy theory land.

 

[leman]

I'm talking about the initial spike lasting seconds, to 80-90W, where temp stays magically flat at 100C and starts dropping only after the power crosses back 50-40W on its way down. Only after that it bounces back and stabilizes at 100C around 45W.

Ah, I understand now. But it's also where the fans are ramping up and the cooling system has not reached an equilibrium yet. There is probably also some inertia to how the heat spreads. Well, we cannily speculate I guess...
[doublepost=1550501658][/doublepost]

I have hit 97/98c frequently, but not 100c before the fans really kick in and it will hover around 96/97c when pushed for an extended period.

Ah, I see. Yes, the laptop I've seen were also moving around 97-99 etc. range. I swipe it all under "100C" since I don't think that the difference of 1-2 degrees is principal at this point. Sorry if this might have lead to a misunderstanding.

But the previous one I did return, used to hit 100c frequently doing even doing random nonsense tasks like syncing my phone via iTunes or opening mail. I would conclude this one is working correctly.

Yeah, that one sounds broken...

 

[MacBH928]

If you're talking about a 2015 15" MacBook Pro, then I'd highly suggest that you open it, give it a good cleaning, and replace the thermal paste. That would do wonders to older MacBook Pros

I think these MBP were made not to be opened and thermal paste is too early for a 4 year old model. Its definitely not the thermal paste, its the hardware cooling and processes handling.

 

[SDColorado]

Back to the OP's dilemma. Does anyone know anything about that mixer.com? I had one of those streams (Robeyonekenobi "Anthem with the Boys) running for about 30 mins while out shoveling snow and when I returned the machine was still at 55c. I tried opening the 4 player split screen into 4 separate windows and enlarging them to split screen. But I am not at all familiar with the site and have no idea what I am supposed to be watching there

 

[Thysanoptera]

Ah, I understand now. But it's also where the fans are ramping up and the cooling system has not reached an equilibrium yet. There is probably also some inertia to how the heat spreads. Well, we cannily speculate I guess...

We know fans can’t get rid of more than 45W of heat, I’ll calculate this, trivial first principle model, just need to grab logs to get proper gradients and this will need to wait until weekend because the MBP is at home while the X1 Carbon is on its maiden trip in my bag. But I’m sure Apple’s cheating, it’s what I do when I need to put a unit I’m working on in a state she’s not supposed to be and not trip it - force a sensor value or limit its reading. But if I leave it in this state after leaving the customer site - I’d better get myself a good lawyer.

 

[leman]

We know fans can’t get rid of more than 45W of heat, I’ll calculate this, trivial first principle model, just need to grab logs to get proper gradients and this will need to wait until weekend because the MBP is at home while the X1 Carbon is on its maiden trip in my bag. But I’m sure Apple’s cheating, it’s what I do when I need to put a unit I’m working on in a state she’s not supposed to be and not trip it - force a sensor value or limit its reading. But if I leave it in this state after leaving the customer site - I’d better get myself a good lawyer.

But isn’t there also a certain inertia to the cooling system? As in, won’t it take a few seconds for heat to spread through the heat pipes/heat sinks during which the CPU could produce more then 45W of power without overheating? That would explain the delayed power reading as well. I’m just speculating here, I don’t have any practical or theoretical experience with thermal engineering...

 

[Webster's Mac]

100C is within Apple's tolerances if the machine is near or at full load. If it's hitting that at only 40% load, you may want to exchange it or take it to the genius bar for a repaste.

 

[Thysanoptera]

But isn’t there also a certain inertia to the cooling system? As in, won’t it take a few seconds for heat to spread through the heat pipes/heat sinks during which the CPU could produce more then 45W of power without overheating? That would explain the delayed power reading as well. I’m just speculating here, I don’t have any practical or theoretical experience with thermal engineering...

Sure thing there is inertia, of course. But inertia doesn’t make the temperature rapidly stop, with 0.01 degree accuracy at some pretermined point. The only physical process that can turn heat exchange process into perfectly isothermic one on the spot in this case would be if the CPU core started to melt. And we’re good 1300C below that. So either that, or Apple somehow got so good at controlling core temp that they can keep it within the 0.01 sensor resolution, but unfortunately only if the power delivery exceeeds the TDP. Otherwise it is a little messy and we’re happy if it stays within 10C, while on launch last year they couldn’t keep it within something like 40C.

Or they cut sensor reading to prevent CPU from emergency shutdown. Take your pick.

 

[casperes1996]

which MBP you have? How long have been using it? I will be amazed to know that at 100C for hours at a time won't affect the internals

I've had a few at this point. None died due to CPU issues. My current is a 2014 base model 15".

note I can tell the temps of the CPU because I use software called iStat Menu... i know of no other way to monitor the CPU heat

There's also MacsFanControl and the older iStat Pro (may not read all sensors since it's old an unsupported, but it's free and lives in Dashboard which I like)

I think these MBP were made not to be opened and thermal paste is too early for a 4 year old model. Its definitely not the thermal paste, its the hardware cooling and processes handling.

Well, these things are all connected. The better the thermal conductivity of the paste, the better the fan will be able to cool it regardless of the quality of the fan.

Or they cut sensor reading to prevent CPU from emergency shutdown. Take your pick.

100C is not the point of emergency shutdown though. It's the point of the CPU slowing down. That slowdown would itself keep the temperatures in check, right? (I think you're more knowledgable about heat transfer than I am, but I can't imagine Apple capping the sensor readings and no class action lawsuit being in the works over it yet)

 

[Eason85]

I have the 8850/4k X1 Extreme. My temps are in the low 40s doing normal work, surfing. Streaming they perk up to the 50s and playing Fallout 4 (the only game I'm playing on it) push the temps to the upper 70s I've yet to see it it 100c.

Quick question : are you running in performance or "smart" cooling mode? After a few hours of gaming with Apex or KF2 I see CPU temps peak at 92, usually at 3.6GHz. The fan likes to run quite slowly though, around 3000 rpm, rather than 5k+.

 

[Thysanoptera]

100C is not the point of emergency shutdown though. It's the point of the CPU slowing down. That slowdown would itself keep the temperatures in check, right? (I think you're more knowledgable about heat transfer than I am, but I can't imagine Apple capping the sensor readings and no class action lawsuit being in the works over it yet)

Yep, you don’t want to hit the wall, so there has to be some buffer after 100C. The slowdown will reduce temps, but when you start down clocking when temps reach 100C you’re already too late. And suing someone for reduced range on a sensor does not make sense, there is nothing malicious about it in itself, it is actually pretty normal.

I do not discount however a possibility that they did it on purpose to get higher geek bench score, or to differentiate MBP from similar Windows laptops who have hard 45W limit. And if this is the case, and it will lead to issues with CPU a year or two from now - now that would be an ambulance chaser equivalent of drunk Walmart truck driver who got wasted in the warehouse together with the management hitting pregnant mother of 4 who was coming back home from charity event which she organized in a local church.

 

[Howard2k]

It's almost as if you'd need to be somewhat cognizant of both the CPU temperature and the system temperature.

(The CPU is obviously fine at 100C, but the system overall is obviously not fine at 100C throughout).

 

[chrizzz09]

TBH I always see people mentioning their fear on the absurd heat on the MBPs and how its not good for the machine, yet I've NEVER seen a post or news from ANYONE actually experiencing a heat related death of their MBP or other machine (and this forum is very notorious when it comes to issues concerning apple products).

Wouldn't there be like thousands of threads if heat damage/heat death would really be a problem or concern? I mean we have already issues like failing keyboards, kernel panics, crackling speakers, stage light ... why no heat related issue?

 

[AndyMacAndMic]

TBH I always see people mentioning their fear on the absurd heat on the MBPs and how its not good for the machine, yet I've NEVER seen a post or news from ANYONE actually experiencing a heat related death of their MBP or other machine (and this forum is very notorious when it comes to issues concerning apple products).

Wouldn't there be like thousands of threads if heat damage/heat death would really be a problem or concern? I mean we have already issues like failing keyboards, kernel panics, crackling speakers, stage light ... why no heat related issue?

Here is Louis Rossman about the temperature on MacBooks. You may or may not like him, but he and his coworkers repair dozens of MacBooks a week. In this video he also mentions several times that he gets a lot of MacBooks with a 'dead' cpu and/or gpu due to overheating:

 

[maflynn]

Quick question : are you running in performance or "smart" cooling mode? After a few hours of gaming with Apex or KF2 I see CPU temps peak at 92, usually at 3.6GHz. The fan likes to run quite slowly though, around 3000 rpm, rather than 5k+.

Mostly on quiet but I also have the performance set to better battery plus I undervolt and I think FO4 is probably less demanding.

 

[leman]

Here is Louis Rossman about the temperature on MacBooks. You may or may not like him, but he (and his coworkers) repair dozens of MacBooks a week. In this video he also mentions several times that he gets a lot of MacBooks with a 'dead' cpu and/or gpu due to overheating

Disregarding the fact that Rossman is a manipulative and dishonest person, I wouldn't be surprised if that particular bit was true. There are not many shops that offer affordable component-level repairs for laptops, so its not strange if many cases they get are repairs deemed impossible or uneconomical by other shops.

 

[mrex]

Disregarding the fact that Rossman is a manipulative and dishonest person,

what??

 

[mrex]

People are amazed by many things. That doesn't make it true. If the internals were at risk by 100C heat, then Apple engineers would have built the machine to keep the temperature lower.

best of the week (had to take a screenshot to save this!)