Yes, your MacBook Pro is running too hot, and yes, it is Apple's fault.

[maflynn]

The difference is Honda's cars are cheap, whereas Apple's computers are expensive.

Have you looked at the price of cars, there's no way you can say they're cheap.

Besides you're missing my point. Buying second hand and/or being out of warranty means that its not the manufacturer's responsibility to fix it. This is not just for computers (or cars) but everything that is made. The argument can be made for TVs, buy a used TV off craigslist that is out of warranty, there's no way the manufacturer should be or would be responsible for any repairs.

 

[jesaja]

Staying with the car analogy, I have to say I agree with the OP.

If I bought a car that, after driving it full speed on the Autobahn for an hour, would throttle itself to 60mph because it overheated, I'd have them fix it or take it back immediately. And so would everybody else!

If they put in an engine revving to 7000rpm, they better make sure the cooling system is up to the task.

Same SHOULD be true for computers.

 

[ColdCase]

Thats like apples and oranges.

Its been standard practice for years to throttle back non critical (non life support) processing when equipment is overheating. For life support, the users take the risk, and then replace the computer post mission as reliability is compromised. All the server farms in the world will throttle back on performance when temperatures rise.

The state of the art in cramming components into a light weight, small size, low power, portable environment while providing max performance 24/7 is not there yet.... at least in the affordable consumer space.

The alternative is to always run low speed.... but apple and other vendors provide the capability to push the envelope of performance while adding enough smart technology to pull things back when conditions warrant.

Another example of the state of art, many consumer SSDs throttle data throughput in order to reach the devices warranty lifetime.

Just saying that there seems to be a failure to understand technology here, not saying that apple could have chosen different parameters.. and other compromises... and fair game for discussion, but car analogies, come on...

You did not pay for nor does apple advertise 100% performance 24/7 regardless of environment. If you need that you need to look at industrial/military device$.

 

[GGJstudios]

I just find it a little twisted that you consider a $2000 professional laptop that throttles at CPU usage "user error." That's usually "engineer error", or "designer error"... not user error.

If your complaint is about the CPU throttling at high temps, talk to Intel. Apple doesn't throttle their CPUs at any temp. That's part of Intel's design, since Intel is the CPU manufacturer.

That's something they don't give you on the specs page at most retailers. Users do not have the information they need to make an informed decision.

The vast majority of users never push their computers hard enough to experience throttling, so it's not a factor in their buying decision. As I've said before, for those who demand more from a computer, they should do their research before buying. This forum has countless threads that show examples of people doing that research, where prospective buyers ask detailed questions about how well a particular Mac model will hold up to playing games, encoding videos, etc. Some get into detailed discussions about CPUs and GPUs. They're smart enough to know what they need and smart enough to ask questions before buying, to make sure the computer they buy will meet those needs. Yes, users do have the information they need, if they're willing to look for it.

Saying that you expect people to dig through numerous web forums to find a bug just makes it sound like you will defend Apple to the death, even if they are wrong.

No, I'm saying that for the vast majority of the millions of Mac buyers, Apple has given them all the information they need to compare their products to competitive products. Apple's competitors aren't advertising throttling specs on their computers, either. So why aren't you blaming them? It sounds like some will attack Apple to the death, even if Apple is right.

 

[Exana]

If your complaint is about the CPU throttling at high temps, talk to Intel. Apple doesn't throttle their CPUs at any temp. That's part of Intel's design, since Intel is the CPU manufacturer.

Before posting such statement, read Intel PDF about cooling for OEM and ODM. Intel explain all requirement for cooling. Throttle only happens when cooling system is not enough efficient. Throttle is also far better than CPU shut down for safety reason. When CPU throttle, you do not loose your stuff with a computer reboot. Intel do not tell to Apple witch CPU they to use.
If Apple wants to offer laptop with most powerful Quad Core, they have to build a great cooling device. Otherwise, they can use a low TDP Quad Core or only Dual Core. But Apple prefers to act has any PC manufacturer : biggest CPU and highest frequencies to seduce.

Why put a Quad Core if it's impossible to use it at is best ? Juste place a Dual Core i7 witch can run at full speed with less heat and battery life.

 

[l.a.rossmann]

This is not just an Apple thing. This is an industry thing and they can't stay in business if their equipment lasts forever. They can't make it if every user keeps their system for five+ years.

Note: I'm not saying MBPs won't last longer than three years of CPU-crunching use. Just saying they penny-pinch wherever they can enough to safely fit their profit models.

Lenovo profited almost half a billion last year making computers that run cooler than Apple that will last 5 years even if you throw them at a wall or spill water in them.

----------

If your complaint is about the CPU throttling at high temps, talk to Intel. Apple doesn't throttle their CPUs at any temp. That's part of Intel's design, since Intel is the CPU manufacturer.

Yes, Intel does have their CPUs throttle. That is the point. The CPU throttles to keep it from being killed when it is run in an abusive environment. The 2011 MBP is an abusive environment. It should not be for $2000.

Apple computers can't be all things to all people. What they should be, at $2000, is properly designed for the chosen processor for all people.

 

[GGJstudios]

Before posting such statement, read Intel PDF about cooling for OEM and ODM.

Yes, I'm quite familiar with that, and I know how throttling works.

Why put a Quad Core if it's impossible to use it at is best ? Juste place a Dual Core i7 witch can run at full speed with less heat and battery life.

It's not impossible to use the quad core, as millions of Mac users demonstrate every day. Some users, however are never satisfied. If Apple offers only dual core, they complain that there's no quad core available. If Apple offers quad core, they complain about heat. If Apple redesigns the MBP to provide much better cooling, they complain it's not thin enough.

The old saying is true: "You can't please all the people, all of the time." So Apple makes the intelligent business decision to satisfy most of the people, most of the time. There will always be those who are never satisfied, no matter what. The fact remains that there IS no heat issue for the vast majority of MBP users and the few who are complaining about it could have bought another computer, better suited to their needs. Moaning about Apple not building computers to satisfy the whims of a small minority of users displays a lack of understanding about what makes a business successful on the scale that Apple operates.

Apple computers can't be all things to all people.

My point, exactly.

What they should be, at $2000, is properly designed for the chosen processor for all people.

You contradict yourself.

 

[l.a.rossmann]

As I've said before, for those who demand more from a computer, they should do their research before buying. This forum has countless threads that show examples of people doing that research, where prospective buyers ask detailed questions about how well a particular Mac model will hold up to playing games, encoding videos, etc. Some get into detailed discussions about CPUs and GPUs. They're smart enough to know what they need and smart enough to ask questions before buying, to make sure the computer they buy will meet those needs. Yes, users do have the information they need, if they're willing to look for it.

Consumers shouldn't have to dig through forums to research the design flaws of a $2000 product before buying them. They should be beat out of the product in R&D. This is nonsense. People who defend Apple for legitimate flaws make the entire Apple fanbase look ridiculous.

It is also obvious that there was a LOT of time to buy a 2011 MBP BEFORE people found out it throttled under full load.

Maybe we should open a forum where people buy Apple products before launch and test all the flaws and put them in stickies so your point gets across better.

 

[GGJstudios]

Consumers shouldn't have to dig through forums to research the design flaws of a $2000 product before buying them.

It's only a design flaw in your opinion. For millions of users, it works quite well as designed. Also, you keep saying "$2000". It doesn't matter what the price is. People would complain if it was $1000 that it didn't perform like a $3000 computer. Basically, you're demanding extreme desktop performance from a notebook, rather than simply buying a desktop computer. It makes no sense at all.

People who defend Apple for legitimate flaws make the entire Apple fanbase look ridiculous.

Again, what you claim is a flaw isn't at all for most users.

 

[l.a.rossmann]

You contradict yourself.

There is a difference between a computer being everything to everyone, and a computer being what it is SUPPOSED to be to begin with!

A professional laptop marketed to the professional segment for video editing & running complex sessions with a top of the line core i7 processor SHOULD BE ABLE TO RUN AT FULL SPEED.

Can this laptop do what an alienware m17 can? NO, it's not trying to be.. it doesn't have the huge honking video card, nor the space for it.

Can this laptop be as light as an 11" macbook air can? NO, it's not trying to be, it cannot fit its power inside that tiny space.

But, if it is going to be a CORE I7 laptop with a HIGH END processor - it should be able to do that well.

I am not saying it should be all things to all people. I am saying it should be to itself what it was made to be. If it can't do that, put a low end core i3 processor in there and call it a day.

 

[Misky]

This topic worries me.. I'm about to buy a 13' Macbook Pro mid 2012 non-retina, and I hope to G-d I won't be able to fry an egg on it while I'm watching a Youtube video...

Could it simply be the aluminium?

 

[B...]

This topic worries me.. I'm about to buy a 13' Macbook Pro mid 2012 non-retina, and I hope to G-d I won't be able to fry an egg on it while I'm watching a Youtube video...

Could it simply be the aluminium?

It could be the aluminum, but the temps (for the most part) are pretty ok for Macs. Flash may use CPU and heat it up, but it should not even get to 65 degrees C. HTML5 is even better.

 

[Exana]

Yes, I'm quite familiar with that, and I know how throttling works.

It's not impossible to use the quad core, as millions of Mac users demonstrate every day. Some users, however are never satisfied. If Apple offers only dual core, they complain that there's no quad core available. If Apple offers quad core, they complain about heat. If Apple redesigns the MBP to provide much better cooling, they complain it's not thin enough.

I remember people on Mac forums laughing about too powerful PC laptops that were noisy and freeze due to overheating (some kind of paradoxes). Over specs was supposed to be the way for selling those notebooks. Now, Apple use the same way and it looks normal ?
I don't car my 450 € netbook/notebook 11.6" Asus S200e throttle when doing serious tasks. It's not build for this. The cooling did not even have a heatpipe, it's a aluminum plate with a fan to cool the Core i3 3217U.
When I buy a MacBook Pro Retina 15,4" advertized to be an powerhouse unit for video encoding and so on, it's not supposed to throttle when doing what it is supposed to do flawless. Where I live, it's never really hot. I dodn't have throttling issue. But when traveling in some countries, it happens.
The MacBook Pro Retina cooling is just good enough to suit common use and meet Intel minmal's prescriptions. At this price point, advanced users can a least expect a great cooling system not only a *good enough*.
My Retina 15,4" is really fine but it fails to be perfect just for those HSF detail... Making a mirror surface will perhaps cost 5$ or 10 $ more, Apple will never lost money to make some powerusers (1 to 3% of the buyers ?) happy.

BTW, I still haven't found a better unit than my Retina so I can't really complain but just want to tell it could be perfect with a better cooling system.

 

[m85476585]

A recurring fallacy in this thread is that the CPU temperature is somehow linked to the temperature of the outside case. It turns out the temperature is not really relevant, it's the power output that matters.

Consider these two scenarios: First, we have a CPU putting out 45W inside a MBP, coupled to the heatsink with some thermal resistance. The CPU temperature might reach 105C, and the outside of the case might reach 60C.

The second scenario: There is a tiny 45W light bulb inside the MBP instead of a CPU, with a filament temperature around 3000C, and it is coupled to the same heatsink with some thermal resistance. The thermal resistance in this case must be much higher, or the light bulb will never reach the operating temperature, so we can use an air gap.

In both scenarios, 45 watts of energy is being dissipated in the MBP, and 45 watts has to leave through the case and the heatsink. Since the case, heatsink, and power dissipation are the same, the outside temperature will be *exactly the same* even though one MBP has a 105C processor and the other has a 3000C light bulb filament.

Applying this analogy to the problem in this thread, improving heat transfer from the CPU to the heatsink shouldn't change the case temperature if all else is constant. And taken to the limit with perfect conduction between the CPU and the heatsink/case, the CPU temperature would be the same as the case temperature, but the case temperature would be no different than in the case of poor heat transfer.

In fact, lowering the CPU temperature might cause the computer to decrease the fan speed, which would *increase* the case temperature because less air is moving past the heat sink!

In short, there are some valid complaints:
1. The outside of my MBP gets too hot
2. The fans are too loud

But the CPU temperature is irrelevant unless it is throttling or hitting thermal shutdown (I think that happens around 125C).

As long as the power dissipated internally is constant, the design of the laptop is a trade-off between case temperature and intrusiveness of cooling. Apple could make the heatsinks larger, but that would make the laptop less aesthetically pleasing. Or they could ramp up fan speed, but that would make it noisy. Or they could make it run hot, while being aesthetically pleasing and quiet (most of the time), and that's what they've chosen to do.

The NASA/JPL paper linked to previously shows that the devices they were testing could be expected to run for 10 years at the TJmax. Is anyone here using a laptop from 2003 and running it at 105C for the past 10 years? I didn't think so.

Edit: I should add that the aluminum makes the case feel hotter than another material would. Since aluminum is a very good conductor of heat, it conducts more heat to your skin than plastic or some other insulator. Of course, the aluminum helps with overall heat dissipation, so it replacing it with plastic would mean higher case temperatures and higher fan speeds (even if the case didn't feel as hot).

 

[wonderspark]

When I read this thread, I thought, "Oh, no... my MacBook Pro needs to be torn apart again so I can polish the heatsinks." I never knew about Intel® Power Gadget, so I downloaded and ran it while doing some folding@home on the laptop.

I ran it for several hours on 7 of 8 threads (700% CPU), and was pleased to see that the turbo boost ran up to 3.1GHz @ ~40W. No throttling, and the CPU core PECI is ~209ºF the whole time, fans about 4800rpm.

Then I bumped it up to max folding for a few hours, and get about 798% CPU. Fans about 5600rpm, temp of 210ºF, and Power Gadget shows the same 3.10 GHz @ ~41.4W. Still, no throttling.

When I'm just using the laptop for web surfing or whatever, the temps are down around room temperature up to 100ºF, fans always at their 1995rpm minimums. I was interested to see the CPU Freq. drop down to about 1.25GHz, though. I guess it slows way, way down when it doesn't need to work. I've seen it as low as 1.2GHz, even. Fascinating!

When I edit 1080p video on this laptop, it plays smoothly and keeps me happy so far. I'll have to run the Power Gadget during my next edit and see where the CPU Freq. goes. I need to find something that gets closer to that black line at 45W.

My conclusion is that this laptop does not throttle the CPU when pushed to the max for hours, unless by throttle, you mean it doesn't reach the max turbo boost of 3.3GHz, only hitting 3.1GHz. (It's the 2.3GHz CPU.)

I'm very happy with what I've learned about my laptop, and that I don't have to worry about pulling it apart again for any heatsink polish/Arctic Silver 5 work. It seems happy to run for days on end at 209-210ºF.

Thank you for this thread, and maybe, when I end up swapping one of the SSDs for the new M500 960GB SSD, I might take the time to polish the heatsinks... but I certainly won't open it up just to do that alone.

 

[Queen6]

launch Terminal, and Activity Monitor so you can easily observe the CPU load and system resources.

To stress test your Mac, type the following command into Terminal:

yes > /dev/null &

This sends one instance of ‘yes’ into the background, to fully load the CPU you need to repeat the process by hitting up arrow and return to run another several instances, or throw a group onto a single line like so:

yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null & yes > /dev/null &

When finished, in the same terminal window type “killall yes” into the command line to kill all instances of the yes command, you will see the following:

$ killall yes
[1] Terminated: 15 yes > /dev/null
[2] Terminated: 15 yes > /dev/null
[3] Terminated: 15 yes > /dev/null
[4] Terminated: 15 yes > /dev/null
[5] Terminated: 15 yes > /dev/null
[6] Terminated: 15 yes > /dev/null
[7]- Terminated: 15 yes > /dev/null
[8]+ Terminated: 15 yes > /dev/null

All instances of “yes” will drop from the process list in Activity Monitor. This will max out the systems CPU, if you want to really want to push the system, have something graphically intensive running in the background triggering the discrete GPU. Just be warned by doing this system core temp`s will be close to or exceed 100C/212F. My Late 2011 2.4 15" MBP CPU peak`s at over 57W (Intel Power Monitor), My Mid 2012 2.3 Quad Core Retina has far better thermal management and easily outperforms the 2011 MBP.

 

[andymac2210]

The amount of cognitive dissonance in this thread is really mind blowing.


When confronted with cold, hard facts from the OP, people just make every excuse under the sun, some of them more hilarious than others.

"It's ok for the vast majority of users" etc.

I think it comes down to two facts.

1. Apple uses headsink components that are comparable to something you'd find in a low end, $300 netbook.

2. This is unacceptable for a $2000 laptop(I won't say professional laptop because then there'll be an argument about the Macbook 'Pro' moniker).


I mean, come on folks, the fact they could simple use a better headsink, a rudimentary component that requires very little engineering instead of doing wild things with air extraction and asymmetrical fans to reduce temperatures by a significant amount tends to suggest they SHOULD do it.

There's NO EXCUSE you can make and you all know it.

 

[JronMasteR]

The amount of cognitive dissonance in this thread is really mind blowing.


When confronted with cold, hard facts from the OP, people just make every excuse under the sun, some of them more hilarious than others.

"It's ok for the vast majority of users" etc.

I think it comes down to two facts.

1. Apple uses headsink components that are comparable to something you'd find in a low end, $300 netbook.

2. This is unacceptable for a $2000 laptop(I won't say professional laptop because then there'll be an argument about the Macbook 'Pro' moniker).


I mean, come on folks, the fact they could simple use a better headsink, a rudimentary component that requires very little engineering instead of doing wild things with air extraction and asymmetrical fans to reduce temperatures by a significant amount tends to suggest they SHOULD do it.

There's NO EXCUSE you can make and you all know it.

So true. I mean, look how they apply the thermal paste... I mean this is horrible and there is no improvement in sight. I mean how difficult can it be to use like a tool which applies the perfect amount of paste?
I paid 4k for mine here in Switzerland, and this is disappointing.
For their top of the line Notebook they should use quality thermal paste and a high quality finish on the heat sink.
As far as I know, the base model CPU doesn't get as hot as the 2.7 or 2.8ghz, about 85-90°, but my 2.7ghz goes up to 103°, and it does throttle.
So the customer is not getting what he paid for because of cheap and badly applied thermal paste and a low quality heat sink.

 

[Doward]

When I read this thread, I thought, "Oh, no... my MacBook Pro needs to be torn apart again so I can polish the heatsinks." I never knew about Intel® Power Gadget, so I downloaded and ran it while doing some folding@home on the laptop.

Agreed, that's why I posted the items - individuals can simply check to see if their system is one of the ones out there throttling themselves. If so, I've provided clear information on how to stop that. It's like a CPU upgrade for 45 minutes of your time

My Late 2011 2.4 15" MBP CPU peak`s at over 57W (Intel Power Monitor), My Mid 2012 2.3 Quad Core Retina has far better thermal management and easily outperforms the 2011 MBP.

I've clocked a high of 61W. I haven't looked into a way to log the output yet. Would be handy info, though.

The amount of cognitive dissonance in this thread is really mind blowing.


When confronted with cold, hard facts from the OP, people just make every excuse under the sun, some of them more hilarious than others.
...

There's NO EXCUSE you can make and you all know it.

Interesting, but I had a good idea of what would happen when I posted this. I hoped to get actual signal instead of all the noise, but the stereotype of the Apple community at large was proven to have some basis of truth after all.

On the flip side, some very good conversation has also come out of this, so I'm pleased that I'm not the only one taking note.

As far as I know, the base model CPU doesn't get as hot as the 2.7 or 2.8ghz, about 85-90°, but my 2.7ghz goes up to 103°, and it does throttle.
So the customer is not getting what he paid for because of cheap and badly applied thermal paste and a low quality heat sink.

I have issue with a system forced to throttle itself to keep from overheating

It hasn't been pointed out, but something to consider for those that claim "it'll shut down if it gets too hot" - you all do realize that is AFTER it has throttled itself down, right? Unacceptable, in my book.

I'm still compiling data on this project.

 

[JronMasteR]

I have issue with a system forced to throttle itself to keep from overheating

It hasn't been pointed out, but something to consider for those that claim "it'll shut down if it gets too hot" - you all do realize that is AFTER it has throttled itself down, right? Unacceptable, in my book.

I'm still compiling data on this project.

Yes, the system will almost never shut itself down, unless there is a real defect. First it will throttle itself until the temps go down. The current cooling performance is not good enough to keep my CPU below 105° on full load.
It's pretty obvious, it goes up to 103° pretty quickly and it just stays there. I will reapply the thermal paste and polish the heatsink. I need to clean the dust out anyway.

 

[JohnDoe98]

I'm still compiling data on this project.

Well thanks for sharing your information and posting this thread. Apple likes to claim they are all about making the best possible product and caring about quality over cost. Some will claim this is just marketing speak, just like the label "pro" on a laptop is not defined and can mean anything. Well, that's partially true, but that quite misses the point and there is something disingenuous and perhaps even illegal in false advertising.

But anyway, that's besides my point. I just wanted to echo that, as you mentioned, the increased costs to optimize these systems further would be rather small/insignificant. I'd like Apple to live up to their standards claiming to care about even the finest little details in the manufacturing process. I for one would be willing to pay an extra 20$ on my 2000$+ laptops if it means it runner (demonstrably) more efficiently.

Also, hasn't anyone noticed your new laptops always run rather cool and quick when you first get them (relatively speaking), but a few years down the road when all the software is more demanding, that's when you start encountering the heat/throttling issues? H265 will be even more demanding, and will become ubiquitous. A lot of laptops will struggle with it just as they do with flash. If the OP's suggestions can help them run a little smoother, for a little longer, well then the extra care in production is worth the effort.

When my warranty is out on my RMBP, I might try this myself, but that will be a while.

Oh, one more thing. When Ive told us about asymmetrical fan blades or the use of sapphire glass, we all said terrific, even though those might have slightly increased costs. We said, wow, what attention to detail. Now ask yourself, if Ive announced that he has increased the quality of the finish on the heatsinks, resulting in lower temperatures, would you praise him or respond to him him as you did the OP? My guess is its the manner the OP expressed himself rather than the message you have an issue with.

 

[fatlardo]

Anyone have any updates on this or more info to add for data?

 

[Xgm541]

I had a custom built iBuyPower.com laptop. Horrendous cooling, processor damaged and couldnt use without throttling the clocks manually.

I had an ASUS laptop that would overheat constantly running mundane tasks. Computer was cleaned from dust/debris, etc.

I had a dell netbook, one of those intel atom ones. No throttling, nothing. I dont think that thing even had a fan in it.

I had a DELL laptop, an XPS1330 (their PRO line) which had horrible cooling for the GPU.

I had a macbook air 11 inch, 2010 edition. Even at 100% CPU/GPU utilization, it worked fine. Yes it was warm, but no throttling.

I now have a 15 inch retina. I play games all the time, watch flash videos, etc. No throttling, no slowdowns.

Also to compare to the car example - if i buy a bugatti veyron I get 10MPG. Im paying premium price, but that's not premium mileage. I might as well go buy a smart car that gets well over 100mpg, right?

Also I'd like to see someone drive on the autobahn in any fast paced car at full speed (RPM) for an extended period of time without your car overheating.

 

[Rossatron]

Do you see how reflective the surface is? That means there are very few valleys. I'd say that Swiftech heatsink has an RA of 3-5 microinches.

When I pulled the heatsink out of my Macbook Pro, I was astounded. While I don't have a profilometer available, I'd estimate the RA of the heatsink copper pads in the neighborhood of 40-60 RA. That's freaking pathetic in a system that you pay the kind of money you do for a quality Macbook Pro. I've seen better heatsink bases in a DELL for crying out loud. Now, my lapping was not the best job in the world. I may have gotten the RA down to the 25 range, but it certainly helped.

How do you expect us to believe you if you are basing your entire theory on "estimate", while not having the capability nor equipment to actually measure things?

 

[Doward]

How do you expect us to believe you if you are basing your entire theory on "estimate", while not having the capability nor equipment to actually measure things?

Do you need a measuring device to tell you if something is closer to a centimeter or a meter?

I've done two more systems (13" 2009 1.86Ghz C2D and a 15" 2008 2.4Ghz C2D) - had an average temperature drop of 6C on the 13", and 11C on the 15".

I'm still working on more data, don't worry.