Resolved rMBP at 187 Degrees F and 2500RPM Fan

[niblet]

I just bought my rMBP last week, and it always has overheated, being very hot to the touch, but the fan would always kick in after a minute or two of the intense heat. I started playing Minecraft today, and the computer shot up to nearly 190 degrees F, and the fans remained at around 2500 RPM even though the computer was almost burning hot to the touch. I tried using SMC fan control, but even with the fans on around 4000 RPM the temperature of the computer stayed at around 170 degrees. I am using Temperature Monitor to monitor the temperature, and am wondering if it is normal for the computer to get this hot without increasing fan speed. The heat especially stinks since even the WASD keys are getting hot to the touch. So, should I take my computer to the Genius Bar to get checked out, or is this normal?

The temperature is totally normal, if you are talking about CPU temperatures. That's about 86 degrees C, which is well within tolerances and substantially below the max it should reach.

I'm a little surprised about you describing it as "burning hot," though. Even with gaming, my rMBP never gets that hot (unlike previous models). So, that may be a reason to have them look at it, although I suspect they won't do anything.

 

[john123]

I just bought my rMBP last week, and it always has overheated, being very hot to the touch, but the fan would always kick in after a minute or two of the intense heat. I started playing Minecraft today, and the computer shot up to nearly 190 degrees F, and the fans remained at around 2500 RPM even though the computer was almost burning hot to the touch. I tried using SMC fan control, but even with the fans on around 4000 RPM the temperature of the computer stayed at around 170 degrees. I am using Temperature Monitor to monitor the temperature, and am wondering if it is normal for the computer to get this hot without increasing fan speed. The heat especially stinks since even the WASD keys are getting hot to the touch. So, should I take my computer to the Genius Bar to get checked out, or is this normal?

The temperature is totally normal, if you are talking about CPU temperatures. That's about 86 degrees C, which is well within tolerances and substantially below the max it should reach.

I'm a little surprised about you describing it as "burning hot," though. Even with gaming, my rMBP never gets that hot (unlike previous models). So, that may be a reason to have them look at it, although I suspect they won't do anything.

 

[niblet]

Thanks, I may have exaggerated a bit about the "burning hot" thing, but it is uncomfortably warm especially when using WASD keys whilst gaming. Thanks for telling me that this is normal, because I really have nothing to compare it to except a 2011 (I think) Dell Inspiron with first or second generation Intel HD graphics and an i3 processor. Thanks!

----------

One more question, what is the max operating temperature for the CPU and GPU of the rMBP 2012

 

[john123]

One more question, what is the max operating temperature for the CPU and GPU of the rMBP 2012

105 degrees C for the 2012/early 2013 CPUs; 100 degrees C for the late 2013 CPUs.

For the 650M, it's also 105 degrees C.

 

[MacSumo]

A 100C+, uncomfortably hot keyboard is normal for a rMBP. Enjoy.

 

[john123]

A 100C+, uncomfortably hot keyboard is normal for a rMBP. Enjoy.

Unfortunately this is a grossly incorrect characterization.

• The CPU is designed to be able to run up to 100 degrees (105 for your model). That's true of all laptops, not just rMBPs.
• "Uncomfortably hot" is not normal. Pre-Retina models definitely were, but cooling has been much improved on Retinas (despite the thinner form factor). Assuming you're not using your laptop on a surface like a pillow, that's where my concern comes in.

 

[niblet]

Unfortunately this is a grossly incorrect characterization.

• The CPU is designed to be able to run up to 100 degrees (105 for your model). That's true of all laptops, not just rMBPs.
• "Uncomfortably hot" is not normal. Pre-Retina models definitely were, but cooling has been much improved on Retinas (despite the thinner form factor). Assuming you're not using your laptop on a surface like a pillow, that's where my concern comes in.

It is only uncomfortably hot when I touch the area behind the keyboard and in front of the hinge, which is where the CPU and GPU is (I think). It is hot to the touch in that area and at times, I cannot leave my hand on it for too long due to the heat. Its not really an issue when I use it, since I never touch that area in normal usage, I am just concerned that my computer has manufacturing defects that could shorten the life of my computer, or potentially cause it to become a fire hazard. Also, I am using my computer on a completely flat desk, and I have only moved it once.

 

[john123]

It is only uncomfortably hot when I touch the area behind the keyboard and in front of the hinge, which is where the CPU and GPU is (I think). It is hot to the touch in that area and at times, I cannot leave my hand on it for too long due to the heat. Its not really an issue when I use it, since I never touch that area in normal usage, I am just concerned that my computer has manufacturing defects that could shorten the life of my computer, or potentially cause it to become a fire hazard. Also, I am using my computer on a completely flat desk, and I have only moved it once.

Oh. Well, those areas are supposed to be hot. That's where the heat's dissipating. I just put my hand there after ramping things up, and it's pretty toasty too.

From what you've described, it doesn't sound like you have a manufacturing defect. However, I do go back to what you said though about the keyboard being hot to the touch. I've slammed my CPU and GPU on my rMBP (actually, 3 models of them), and I've never found the keyboard to be anything other than warm. Obviously I can't compare mine to yours. If you're worried, I see no harm in taking it to the Genius Bar and asking. The only cost is, of course, your time. All that said, the temperatures you report are A-OK.

Lastly, I'll note that some folks here have an axe to grind for some reason, and contend that there's a cooling design problem with these laptops. However, the truth of the matter is that heat is just normal when pushing your CPU or GPU. Apple could have done some things to improve cooling, but that doesn't mean that the status quo production is problematic. So in short, don't let the alarmists get you too worried.

 

[niblet]

Thanks, john123, the more that I think about it, the keyboard was not really hot, but more so warm like you said. Thanks, and I apologize for taking up so much of your time.

Thanks-Really starting to love this whole forum thing

 

[john123]

Thanks, john123, the more that I think about it, the keyboard was not really hot, but more so warm like you said. Thanks, and I apologize for taking up so much of your time.

Thanks-Really starting to love this whole forum thing

Don't apologize—that's what everyone is here for! Congrats on your purchase, and enjoy it!

 

[richard371]

I just got the new 15 rMBP wGPU and it so hot just doing a few task like updating windows in a VM. This is the biggest POS I Have spent 2599 on in my life. I wanted the extra 8GB ram and SSD etc for my VMs. I was going to return the rMBP 8/256 for this but its about to explode where the 13' never even gets warm.

Also the screen has a more yellow tint then my 13 esp on the right hand side. This machine is a complete disaster. Im really disappointed in apple.

 

[Doward]

105 degrees C for the 2012/early 2013 CPUs; 100 degrees C for the late 2013 CPUs.

For the 650M, it's also 105 degrees C.

Incorrect - that is the absolute maximum temperature recommended by the manufacturer (Intel / Nvidia).

No engineer runs on the limits like that, and Apple's no different. Apple's thermal profiling system is designed to keep the CPU/GPU temperatures under 95C.

At no point show anyone have stable loaded temperatures exceeding 95C, and if you manage 90C or below, your thermal system is working perfectly.

 

[john123]

Incorrect - that is the absolute maximum temperature recommended by the manufacturer (Intel / Nvidia).

No engineer runs on the limits like that, and Apple's no different. Apple's thermal profiling system is designed to keep the CPU/GPU temperatures under 95C.

At no point show anyone have stable loaded temperatures exceeding 95C, and if you manage 90C or below, your thermal system is working perfectly.

Sorry, sir, but it is you are incorrect. Yes, those are the maximum temps (Tjunction values), but the notion that the CPU shouldn't approach those limits when being pushed is simply false. That's simply going to happen if you, say, kick off 8 threads that push the CPU as hard as possible. You're under the misguided impression that reaching that value inherently is problematic; you won't find anything from Intel saying that.

Want empirical evidence? Check out theSeb's excellent work in the thread he created: https://forums.macrumors.com/threads/1688255/

 

[MacSumo]

I just got the new 15 rMBP wGPU and it so hot just doing a few task like updating windows in a VM. This is the biggest POS I Have spent 2599 on in my life. I wanted the extra 8GB ram and SSD etc for my VMs. I was going to return the rMBP 8/256 for this but its about to explode where the 13' never even gets warm.

Also the screen has a more yellow tint then my 13 esp on the right hand side. This machine is a complete disaster. Im really disappointed in apple.

Good thing you joined in 2008, or else John would accuse me of duplicate accounts.

 

[john123]

Good thing you joined in 2008, or else John would accuse me of duplicate accounts.

All I ever said was that it was odd that you referred to yourself in the third person in the RAM thread. Because, well, it is.

People can have subjectively different opinions about products. That's fine. What isn't fine is ignoring the data, or twisting the interpretation of that data to support an invalid conclusion—which is what happened in theSeb's excellent thread.

I don't know what richard371 considers "so hot," nor do I know how yellow his screen is. I do know saying it's "about to explode" is hyperbole. And I do know there's not a darn thing wrong with an Ivy Bridge CPU hovering around 100-105 degrees Celsius when under load. Because physics.

 

[Doward]

Sorry, sir, but it is you are incorrect. Yes, those are the maximum temps (Tjunction values), but the notion that the CPU shouldn't approach those limits when being pushed is simply false. That's simply going to happen if you, say, kick off 8 threads that push the CPU as hard as possible. You're under the misguided impression that reaching that value inherently is problematic; you won't find anything from Intel saying that.

Want empirical evidence? Check out theSeb's excellent work in the thread he created: https://forums.macrumors.com/threads/1688255/

The evidence you present states quite clearly that theSeb's 2.6Ghz system throttled down to 2.5Ghz while encoding a file.

Attached is my own system's log from Intel's Power Gadget during a 30 minute encode of a Christmas video I shot this morning.

My system maintains 100% Turbo Boost (~3.30 ghz) and the thermal log never changed from 100% CPU_LIMIT. My temperatures stabilized @ 194F (90C) with fans @ 5200rpm.

This clearly shows there is room for improvement in the thermal system.

105C is Intel's stated *maximum* temperature - running any system at the limits will reduce the life of that system.

 

[leman]

The evidence you present states quite clearly that theSeb's 2.6Ghz system throttled down to 2.5Ghz while encoding a file.

The average clocks during all individual periods was above 2.6Gzh. Thats why he promised to redo the test with longer-running tests

This clearly shows there is room for improvement in the thermal system.

This is absolutely true, and I doubt that anyone would argue with that. It is a shame that Apple seems to have such poor quality in choosing and applying the thermal compound - after all, its a very small thing that does not require extensive redesign or investment, but does seem to bring noticeable benefits.

105C is Intel's stated *maximum* temperature - running any system at the limits will reduce the life of that system.

Pending a proper study, I can't say that I agree with this statement. I mean, sure - but will this reduction in the lifespan be actually noticeable? I doubt that very much.

 

[Doward]

Pending a proper study, I can't say that I agree with this statement. I mean, sure - but will this reduction in the lifespan be actually noticeable? I doubt that very much.

Interesting point - what constitutes a 'lifespan' for a computer system?

I consider a 10 year span. I have systems from 2003 that still operate fine today. My oldest Mac currently is a G3 system that still chugs along just fine.

For the premium I pay for an Apple product, I don't think that's an unfair expectation. If I wanted to churn through products every other year, I'd go the disposable PC route.

 

[john123]

The evidence you present states quite clearly that theSeb's 2.6Ghz system throttled down to 2.5Ghz while encoding a file.

Attached is my own system's log from Intel's Power Gadget during a 30 minute encode of a Christmas video I shot this morning.

My system maintains 100% Turbo Boost (~3.30 ghz) and the thermal log never changed from 100% CPU_LIMIT. My temperatures stabilized @ 194F (90C) with fans @ 5200rpm.

This clearly shows there is room for improvement in the thermal system.

105C is Intel's stated *maximum* temperature - running any system at the limits will reduce the life of that system.

Your entire argument is predicated upon an assertion (bolded for emphasis) lacking any evidence—that running a CPU at or near the maximum has some disproportionate impact on system lifespan. In general, heat isn't good for electronics, but there's zilch in terms of evidence to suggest that running at the Tjunction is problematic.

Could the system be cooled better? Yup. I've agreed with people who have posted that. The thermal paste in particular is not of good quality and isn't applied well. Does that mean there is a problem? Nope.

Also, as noted above, you mis-read the evidence in the other thread. But the throttling is beside the point on this discussion.

You have conjecture without empirical evidence.

 

[leman]

Interesting point - what constitutes a 'lifespan' for a computer system?

There a re number of components (PSU, storage, capacitors) which are likely to fail before a CPU - maybe even a stressed one. Also, computers get obsolete after 4 years or so. If you are a 'professional user' (whatever that might mean) or any other kind of person that needs performance, a 4 year old machine just won't do it.

I consider a 10 year span. I have systems from 2003 that still operate fine today. My oldest Mac currently is a G3 system that still chugs along just fine.

For the premium I pay for an Apple product, I don't think that's an unfair expectation. If I wanted to churn through products every other year, I'd go the disposable PC route.

Sorry to disappoint you, but all computers are disposable. And there is nothing special in the Apple computers when we talk about components - they use the same CPUs, GPUs, microchips etc. as everyone else. Its not like, say, a food processor or a car where you can prolong the lifetime just by using sturdier (and more expensive) materials.

That said, if you are fine with using a 10 year old computer, then overheating should be the least of your worries, as you will never do anything with it that will push it that far.

 

[Doward]

That said, if you are fine with using a 10 year old computer, then overheating should be the least of your worries, as you will never do anything with it that will push it that far.

Depends on the task. Is a ten year old system my primary? Of course not. That's just being asinine. Yet I do have a Pentium II system that runs my basic print server. Why haven't I upgraded it? No reason, it just works fine.

So many people these days have no clue just how much power is under the hood of their systems. Disappointing.

 

[leman]

Depends on the task. Is a ten year old system my primary? Of course not. That's just being asinine. Yet I do have a Pentium II system that runs my basic print server. Why haven't I upgraded it? No reason, it just works fine.

So many people these days have no clue just how much power is under the hood of their systems. Disappointing.

Of course its perfectly sufficient to run a print server, a personal storage server, even a small webserver! That is still no valid reason to keep it around. You can build a Atom platform for 100 euro that would be just as good at doing those same things - and consume just a fraction of the electricity/take much less space than your Pentium II. In your case: why not get a printer with a LAN port and a built-in server? The power bill savings alone compared to your current setup will probably pay it off in less then half a year. Yes, tinkering with old computers is fun. But in most cases, that's all there is to it. I don't think its a good strategy to make a purchase decision based on your wish to tinker with the computer 10 years in the future.

 

[Doward]

But make a purchase decision based on your wish to tinker with the computer 10 years in the future.

Ah, there's our disconnect.

I'm not advocating that I want a system that will last 10 years precisely because I intend to use it in 10 years. I'm advocating I want a system that will last 10 years because a system that lasts 10+ years must, by definition, be built to a higher quality standard.

I buy a Lexus for the build quality. And yes, the average 10 year old Lexus will still be much nicer, with less problems, than the average 10 year old Kia.

Also note the Lexus has a *much* lower chance of any sort of failure vs the Kia. That said, I expect a premium product such as a MacBook Pro to have similar high-level build quality.

The tests at hand show that the quality could be improved in the MBP.

*edit* Interestingly enough.... I want to keep my rMBP for a minimum of 8 years

Yes, there are people that do it.

 

[MacSumo]

Ah, there's our disconnect.

I'm not advocating that I want a system that will last 10 years precisely because I intend to use it in 10 years. I'm advocating I want a system that will last 10 years because a system that lasts 10+ years must, by definition, be built to a higher quality standard.

I buy a Lexus for the build quality. And yes, the average 10 year old Lexus will still be much nicer, with less problems, than the average 10 year old Kia.

Also note the Lexus has a *much* lower chance of any sort of failure vs the Kia. That said, I expect a premium product such as a MacBook Pro to have similar high-level build quality.

The tests at hand show that the quality could be improved in the MBP.

*edit* Interestingly enough.... I want to keep my rMBP for a minimum of 8 years

Yes, there are people that do it.

Exactly - a $3K computer should be designed with quality in mind, but I found that lacking in the rMBP. The laptop is okay, but not good enough. Apple should either drop their price or improve quality. They can't expect intelligent consumers to pay a price premium for low quality.

 

[john123]

They can't expect intelligent consumers to pay a price premium for low quality.

Actually, it appears that they can: http://www.apple.com/pr/library/2013/10/28Apple-Reports-Fourth-Quarter-Results.html