What's your opinion on the MacBook Pro not being a professional's computer?

[luffytubby]

More fuel to the fire. Most reliable hardware on a three year basis; http://www.statisticbrain.com/laptop-malfunction-rates/


Wow, Asus!!


and more; http://www.squaretrade.com/htm/pdf/SquareTrade_laptop_reliability_1109.pdf

How Reliable is Your Laptop, Really?
That isn't any easy question to answer. Reliability information isn't easy to obtain, and so most consumers don't even put major thought into it, assuming one laptop is about as reliable as the other. Savvy consumers might ask on forums for information, but the replies found are usually unsatisfactory, based off individual experiences rather than statistics.

There actually is a very good source of information about laptop reliability ratings available, however. This source is the Squaretrade reliability report. The report, created by a company called Squaretrade which deals in laptop warranties, is the most accurate source of information currently available about laptop reliability.

Let's take a look at that report and what it says.

Laptop Reliability Is Generally Poor
Overall, laptop reliability ratings leave a lot to be desired. About 20 percent of laptops will fail due to a hardware malfunction in their first three years of use. While most laptops come with a standard one year warranty, the majority of failures occur after the first year. The first year chance of laptop failure is 4.7 percent. This climbs to 12.7 percent in the second year.


When accidental damage is taken into account, the chance of failure within the first three years of use increases to 31 percent. It is not possible to make any comment on the build quality of laptops because of accidental damage, but this does shows that accidental damage is a major factor in overall reliability. This is a factor which usually isn't a major part of a consumer product's reliability. It is rare, for instance, that an HDTV is destroyed by accidental damage, flying Wiimotes included.

If you purchase a laptop today, there is statistically a one in three chance that it will not be with you in three years. However, you can improve your odds by purchasing a laptop from a brand that has a high laptop reliability rating.


Brand Matters
The brand of laptop which you buy has a major effect on the chance that your laptop will fail. The graph below, which is from the Squaretrade report, shows how the laptop brands stack up when it comes to their reliability.

Laptop Reliability Ratings

As you can see, there is a significant difference between the laptop reliability ratings of different brands. You are 40% more likely to experience a failure within three years if your purchase a laptop from the least reliable brand, HP, instead of from the most reliable brand, ASUS. That is a large gap. In fact, Lenovo, Acer, Gateway and HP all performed very particularly poorly. This can be noted by the large gap between Dell and Lenovo. The fact that Acer and Gateway have virtually the same reliability also brings credit to Squaretrade's report - Acer owns Gateway and designs are often shared between the companies.

ASUS and Toshiba deserve serious credit and serious consideration from any laptop buyer. They are virtually tied in the battle for first place. Anyone looking for a laptop that needs to last several years would be best served by purchasing an ASUS or Toshiba laptop.

Price Equals Reliability
An important trend which shows up in the Squaretrade reliability report is a correlation between laptop reliability and laptop price.

Laptop Reliability

According to the Squaretrade report, netbooks have the worst reliability out of all laptops which are currently made. It should be pointed out that this data is more limited than with laptops because netbooks are a newer product. However, 5.8% of netbooks have failed within the first year, which can be compared to 4.7% of laptops. If the were to continue to the third year of use, you could expect to see an average failure chance of about 25% on netbooks (accidents not included), which is 5% more than the failure chance with a normal laptop.

It also seems that "premium laptops" - laptops over $1000 in price - have a lower chance of failure than laptops which are less expensive. There is an 18.1% chance that a premium laptop will fail within three years, which is a little bit lower than the chance of a normal laptop failing. Still, this difference isn't large enough to justifying spending a grand on a laptop if you don't need a premium system.

Conclusions
Above all, the most important conclusion to take with you from this report is the fact that laptop brand does make a difference when it comes to laptop reliability. Several of the brands which were on the bottom of the list - Acer, Gateway and HP - are brands that often offer laptops with very low prices relative to their hardware. This seems to work in terms of sales, because HP and Acer are the number 1 and number 2 computer vendors in the world (and remember, Gateway is also a subsidiary of Acer). Before you leap at a great deal, consider how long you need your laptop to last. If you want it to last three years or more, you may want to go for a slightly more expensive ASUS, Toshiba or Sony laptop instead.

Another observation which can be made from this report is one that Squaretrade probably has no problems with - laptops are not as reliable as other consumer electronics. Does that mean that you should purchase a protection plan on your laptop? Personally, I say no, it doesn't. While laptop reliability ratings are poor, it is still unlikely that your laptop will fail - as long as you purchase from a reliable brand.

Finally, it can be argued that netbooks are less reliable than laptops. This is not surprising when you consider how inexpensive netbooks are. Of course, despite the increased chance of failure, netbooks are probably not worth worrying about. After all, netbooks are extremely cheap, and you'll probably replace one within two or three years anyway.

 

[FrozenDarkness]

More fuel to the fire. Most reliable hardware on a three year basis; http://www.statisticbrain.com/laptop-malfunction-rates/


Wow, Asus!!


and more; http://www.squaretrade.com/htm/pdf/SquareTrade_laptop_reliability_1109.pdf

pretty interesting read although not statistically significant. Squaretrade data self-chooses since it's only pulled from people who decide to purchase squaretrade. Its entirely unclear if its sample size is representative of the general population. This goes for every laptop manufacturer in their study.

in fact, i'd be willing ot bet that any study that samples warranty purchasers will have a higher failure rate than the general population at large.

 

[Brandon263]

As someone who earns money from using my Haswell MacBook Pro, I can honestly say that I have never had it become too hot to touch. It certainly performs much better than my first generation rMBP and is faster and cooler than my last Dell workstation.

 

[bill-p]

Notebookcheck runs every notebook they test under a sustained (several hours) Prime + Furmark load. Many do not break 90 degrees (the big gaming notebooks or AMD notebooks), or if they are breaking 90 degrees its because they are configured to run at max turbo. But the fact remains that furmark + prime is unrealistic.

Did you know those gaming notebooks also throttled in those tests? Here are some notes:

http://www.notebookcheck.net/Review-Alienware-18-Notebook.102566.0.html

The components leave a mixed impression. On the one hand, it is very positive that neither the GPU nor the CPU exceeds 90 °C (194 °F) even under maximum load and level off at around 80 °C (176 °F). However, this is only possible because of the automatic throttling of the components. While the processor dropped to 2.7 GHz (- 300 MHz compared to the nominal clock), the core clock of the GeForce GTX 780M SLI dropped to 614-666 MHz. That is a deficit of 25 and 19%, respectively.

And talking about which, they reported no throttling for the rMBP under the same condition:

http://www.notebookcheck.net/Review-Apple-MacBook-Pro-15-Retina-2-3-GHz-Mid-2012.78959.0.html

A render is not and the OP was specifically stating that he was reaching 95+ degrees on render projects using consumer software.

Again, it's safe to note that different programs report different temperature readings.

And since no one else can replicate what the OP is experiencing, the logical conclusion is that this is not a widespread problem. Unless you have data that points to otherwise.

I agree that running at base clock is not throttling. However it is far from ideal and shame on apple for selling CPU upgrades that are never felt in the real world (and any manufacturer who does this).

Read back above and tell me why it's okay for... say... Alienware (Dell) to throttle but it's not okay for Apple to do the same? And just to note, once again, that no one can replicate what the OP is doing.

Just for a data point, my rMBP doesn't throttle. It runs so well that I can overclock the GPU further and take the hit with increased temperature.

There are multiple threads about a decreasing battery under load on rmbp 15"; simply put the power supply is not large enough. If the power supply is throttling because its in danger of exploding then there is something wrong with the power supply. The power supply should be able to take the load I apply on it with the computer that is sold with it. Apple needs to make the power supply a little larger and a little more powerful. As it is its a poor match with the rmbp 15".

I'm just pointing to a possible explanation as to why some have seen the decreasing battery.

However, there is another explanation given that notebookcheck above has noted that the rMBP they measured never exceeded the maximum power supply: the battery started discharging after a while to prevent itself from overcharging.

That is also a possible cause for an explosion, and it's just not safe.

And personally, again, I have never ever seen this scenario with any of my rMBP. I have left the machine plugged in overnight running an algorithm the whole night... pegging the CPU at 100% (800% under Mac OSX because there are 8 virtual threads). The next morning, there is still 100% on the battery.

That's very representative of a rendering project as well, and if you're saying that you can somehow realistically peg the computer even more than that (more than what notebookcheck does even) to make it throttle and causing the power supply to throttle, then something is wrong with the computer... or your usage (why are you using a laptop for that?)

I think a lot of what the OP is saying can't be verified as truth... and there have been many evidences pointing to otherwise. So... why are you supporting his position?

And just as an aside, have you actually owned and used a rMBP for any period of time and experienced any of the issues here?

 

[john123]

pretty interesting read although not statistically significant. Squaretrade data self-chooses since it's only pulled from people who decide to purchase squaretrade. Its entirely unclear if its sample size is representative of the general population. This goes for every laptop manufacturer in their study.

in fact, i'd be willing ot bet that any study that samples warranty purchasers will have a higher failure rate than the general population at large.

Good points all around. I'd also add that "failure" has a potentially wide definition.

----------

No. There aren't. Nothing but innuendo and internet gossip. Meanwhile plenty of professionals are quite happy with their rMBP's.

His first proof is a forum post. Hardly proof and laughable.

+1 to this.

 

[iKrivetko]

Good points all around. I'd also add that "failure" has a potentially wide definition.

My thought exactly, seeing how people tend to return macbooks that seem to be creaky, or have an unusually loud trackpad click.

 

[cirus]

And I've participated in them and agreed with some. However, your idea of an "informed statement" is based on assumption and conjecture, not facts, and you have blurred the line, and that's my problem.

A fact is that the thermal paste application and materials done by Apple are mediocre at best and probably sub-par. A conjecture, albeit a reasonable one, is that this increases the probability of failure. Is that a "problem"? It all depends on what that probability is.

This is logic and statistics 101. I can't state it any more clearly and simply.

Then again when looking at extreme probabilities, approximations can be made. Does your conjecture agree with observed data and make logical sense? Yes it does. In almost every/any case where inferior materials are used for mechanical devices the quality decreases as material costs are lowered. Again please do not fall into the logical fallacy of 'you can't prove it absolutely so your hypothesis means nothing'. A strong and very good agreement between hypothesis and the data indicate that the hypothesis works most of the time. Most of the world is based on this 99% approximation.

As for approximations, if you are part of a group of 100,000 people and 99,999 will get a prize can you make the approximation that you are going to get a prize? Yes, because it is extremely unlikely for you to be the one person who doesn't (Its not guaranteed but its so exceedingly likely that you can make the assumption, similar to how falling 50 stories is probably going to kill you but there have been cases where people have lived). The same applies to your example. Using poorer materials will generally increase the probability of failure. That much is a given. However, in this specific case it may not apply. However, since the vast majority of the time that case applies, its not wrong to say that you expect that case to apply (unless some other unknown factor is involved). You do not ignore a correlation because its there only 99% of the time.

Did you know those gaming notebooks also throttled in those tests? Here are some notes:

http://www.notebookcheck.net/Review-Alienware-18-Notebook.102566.0.html

And talking about which, they reported no throttling for the rMBP under the same condition:

http://www.notebookcheck.net/Review-Apple-MacBook-Pro-15-Retina-2-3-GHz-Mid-2012.78959.0.html

Did I not paste this exact quote previously? The rmbp throttled like crazy.

Prime 95 and Furmark can extract the maximum performance out of the hardware. When both programs are run simultaneously, the core frequency drops (caused by Prime 95) to 1.2 Ghz. Running Prime 95 alone, we measured temperatures between 92 and 104 degrees Celsius (197.6 and 219.2 degrees Fahrenheit), and the word "Throttling" flashes repeatedly in HWiNFO. Furmark on its own cycles through all available Turbo Boost steps. After about one hour, the temperature settled at a fairly consistent 85 degrees Celsius (185 degrees Fahrenheit) with the fan system running. We never measured above 100 degrees Celsius (212 degrees Fahrenheit) with Furmark. This limitation might be due to the power adapter: in the above scenario, the system would require 86.6 watts - but the power adapter can only supply a maximum of 85 watts.

Clocks were much higher than the rmbp under the max throttle test (2.7 ghz vs 1.2 ghz). The throttling is also due to the bios dell uses which prioritizes temperature over speed. Maintaining 80 degrees under max load (after a spike to ~87 ish) is pretty good. Again one laptop is not indicative of all laptops and the alienware 18 is clearly in a different weight class.

Again, it's safe to note that different programs report different temperature readings.

And since no one else can replicate what the OP is experiencing, the logical conclusion is that this is not a widespread problem. Unless you have data that points to otherwise.

If a program is reporting 46 degrees as said before under load something is clearly wrong. Its not 'oh different programs report different temperature readings' but this program is not reading something right. There was/is a problem with some AMD notebooks where programs such as HWinfo show something like 100 degrees on the CPU at idle, which is clearly wrong as the computer would shut off (amd's processors generally have a lower tjmax value).

Read back above and tell me why it's okay for... say... Alienware (Dell) to throttle but it's not okay for Apple to do the same? And just to note, once again, that no one can replicate what the OP is doing.

Just for a data point, my rMBP doesn't throttle. It runs so well that I can overclock the GPU further and take the hit with increased temperature.

I'm just pointing to a possible explanation as to why some have seen the decreasing battery.

However, there is another explanation given that notebookcheck above has noted that the rMBP they measured never exceeded the maximum power supply: the battery started discharging after a while to prevent itself from overcharging.

That is also a possible cause for an explosion, and it's just not safe.

And personally, again, I have never ever seen this scenario with any of my rMBP. I have left the machine plugged in overnight running an algorithm the whole night... pegging the CPU at 100% (800% under Mac OSX because there are 8 virtual threads). The next morning, there is still 100% on the battery.

That's very representative of a rendering project as well, and if you're saying that you can somehow realistically peg the computer even more than that (more than what notebookcheck does even) to make it throttle and causing the power supply to throttle, then something is wrong with the computer... or your usage (why are you using a laptop for that?)

I think a lot of what the OP is saying can't be verified as truth... and there have been many evidences pointing to otherwise. So... why are you supporting his position?

And just as an aside, have you actually owned and used a rMBP for any period of time and experienced any of the issues here?

Well there is variation between products but when you design something variation needs to be taken into account so that most machines work as intended.

(does 100% CPU mean 100% CPU on full boost? because that does make a difference).

Needless to say the rmbp 15 does not possess adequate cooling. Temperatures over 85 are undesirable and should not be seen in expensive notebooks. I don't care what manufacturer does it it should not be done. The 13" model does not get nearly as hot.

 

[john123]

Then again when looking at extreme probabilities, approximations can be made. Does your conjecture agree with observed data and make logical sense? Yes it does. In almost every/any case where inferior materials are used for mechanical devices the quality decreases as material costs are lowered. Again please do not fall into the logical fallacy of 'you can't prove it absolutely so your hypothesis means nothing'. A strong and very good agreement between hypothesis and the data indicate that the hypothesis works most of the time. Most of the world is based on this 99% approximation.

I didn't say that lack of absolute proof invalidates a hypothesis. But by comparative standards, the leaps you're willing to make are even further than the people who (laughably) believe the evidence suggests that vaccines cause autism. Here, we don't even have strong data or correlations, hence why I don't think what you believe comes remotely close to passing muster. You've decided you believe it's a problem, and you're intransigent in that belief. As a side note, in the 1500s, everywhere people went, nothing seemed curved, so they also believed the earth was flat.

As for approximations, if you are part of a group of 100,000 people and 99,999 will get a prize can you make the approximation that you are going to get a prize? Yes, because it is extremely unlikely for you to be the one person who doesn't (Its not guaranteed but its so exceedingly likely that you can make the assumption, similar to how falling 50 stories is probably going to kill you but there have been cases where people have lived). The same applies to your example.

Actually, it doesn't. Not at all. You flat out made up a toy 99% statistic, but the probabilities are unknown but certainly don't come anywhere close to that. In fact, this falls into the scope of rare event probabilities, which has a whole set of methodology in statistics because the likelihood of Type I and Type II errors is inverted versus the usual cases.

Using poorer materials will generally increase the probability of failure. That much is a given.

We agree. What we don't know is what the change in incidence is. If it goes from, say, 1% to 1.05% (i.e., a 5% increase in failures), I'm not going to get my panties in a wad.* If, however, the failure rate goes from 1% to 3% (i.e., a 300% increase in failures), then I'd argue that's a really big deal.

* Footnote: I have to assume that there's some reason Apple's and most other big PC company's thermal paste and its application suck, and cost is probably a key driver. I'd rather avoid the philosophical debate about whether Apple should do everything under the sun to minimize failure rates, even at their own cost, or what that threshold can be. For the sake of argument, let's please not get stuck on the numbers and just agree that there's some small amount that isn't really a "problem" and some larger amount that really is a "problem."

If a program is reporting 46 degrees as said before under load something is clearly wrong. Its not 'oh different programs report different temperature readings' but this program is not reading something right.

Again, that was the heat sink. The heat sink is obviously not the CPU. No one in this thread has yet explained where the CPU core temperatures are coming from in some of these tests (whether that's Intel circuitry or something else), and until that's resolved, cross-manufacturer comparisons of temperature readings don't make much sense. The fact that I can't even get Mac programs that are supposed to read directly on the CPU itself to agree on temperatures (I saw a 15 degree band across two different programs) is, well, problematic. There's max and rolling averages, there's max for each core, etc. Apples to apples comparisons are important if we're going to make these kinds of arguments.

Needless to say the rmbp 15 does not possess adequate cooling. Temperatures over 85 are undesirable and should not be seen in expensive notebooks. I don't care what manufacturer does it it should not be done. The 13" model does not get nearly as hot.

And per the above, your first sentence here remains a claim without a warrant. As for the 13" model, it's not a quad core. That in and of itself is super significant. Once again, valid comparisons are the hallmark of valid points.

 

[bill-p]

Clocks were much higher than the rmbp under the max throttle test (2.7 ghz vs 1.2 ghz). The throttling is also due to the bios dell uses which prioritizes temperature over speed. Maintaining 80 degrees under max load (after a spike to ~87 ish) is pretty good. Again one laptop is not indicative of all laptops and the alienware 18 is clearly in a different weight class.

To be clear, I could not replicate notebookcheck's test. (see below for proof)

So either 1.2GHz is a glitch, or they (notebookcheck) have a bad lemon... like the OP does.

If a program is reporting 46 degrees as said before under load something is clearly wrong. Its not 'oh different programs report different temperature readings' but this program is not reading something right. There was/is a problem with some AMD notebooks where programs such as HWinfo show something like 100 degrees on the CPU at idle, which is clearly wrong as the computer would shut off (amd's processors generally have a lower tjmax value).

And I have proof below that temperature reported in a computer program is not reliable.

Because it's reporting 105C, and that would have caused the machine to shut down.

Well there is variation between products but when you design something variation needs to be taken into account so that most machines work as intended.

And they do. Either that or... I have a super Retina Macbook... that's far superior to everyone else's.

(does 100% CPU mean 100% CPU on full boost? because that does make a difference).

Yes.

Needless to say the rmbp 15 does not possess adequate cooling. Temperatures over 85 are undesirable and should not be seen in expensive notebooks. I don't care what manufacturer does it it should not be done. The 13" model does not get nearly as hot.

Please look at the above and tell me it doesn't have adequate cooling?

Just as an aside, the rMBP 13" does get as hot (CPU temperature reaching over 85C).

http://www.notebookcheck.net/Review-Apple-MacBook-Pro-Retina-13-Late-2013-Notebook.105035.0.html

The internal temperatures also show the different fan controls of Windows 8.1 and OS X Mavericks. While the CPU only reached up to 85 °C (185 °F) in our stress test under Windows, we could determine temperatures of up to 93 °C (199.4 °F) for the Core i5 with OS X, especially during gaming.

What can I say? I know my machine is working as intended (see screenshot above), and that's more than enough for me.

 

[john123]

Please look at the above and tell me it doesn't have adequate cooling?

Emphasis added. I'm starting to think this whole thing boils down to you and me saying, "It's adequate and working as intended," and cirus, the OP, and others essentially saying, "But it could be manufactured to run cooler/support more TurboBoosting under existing temperatures." If that's a correct assessment, then maybe there's room for agreement (other than the subjective claims as to whether there is, or is not, a "problem").

 

[cirus]

I didn't say that lack of absolute proof invalidates a hypothesis. But by comparative standards, the leaps you're willing to make are even further than the people who (laughably) believe the evidence suggests that vaccines cause autism. Here, we don't even have strong data or correlations, hence why I don't think what you believe comes remotely close to passing muster. You've decided you believe it's a problem, and you're intransigent in that belief. As a side note, in the 1500s, everywhere people went, nothing seemed curved, so they also believed the earth was flat.


Actually, it doesn't. Not at all. You flat out made up a toy 99% statistic, but the probabilities are unknown but certainly don't come anywhere close to that. In fact, this falls into the scope of rare event probabilities, which has a whole set of methodology in statistics because the likelihood of Type I and Type II errors is inverted versus the usual cases.

My 'leap' if that is what you want to call it is minor and you yourself agree that higher temperatures are non-ideal and lead to wear and tear. What we disagree on is the impact and significance.

The 99% is simply for illustrative purposes. But most scientific theory, most everything we do is based on really got but not 100% accurate estimations.

We agree. What we don't know is what the change in incidence is. If it goes from, say, 1% to 1.05% (i.e., a 5% increase in failures), I'm not going to get my panties in a wad.* If, however, the failure rate goes from 1% to 3% (i.e., a 300% increase in failures), then I'd argue that's a really big deal.

* Footnote: I have to assume that there's some reason Apple's and most other big PC company's thermal paste and its application suck, and cost is probably a key driver. I'd rather avoid the philosophical debate about whether Apple should do everything under the sun to minimize failure rates, even at their own cost, or what that threshold can be. For the sake of argument, let's please not get stuck on the numbers and just agree that there's some small amount that isn't really a "problem" and some larger amount that really is a "problem."

The impact of 95-100 degrees for significant amounts of time (heavy rendering) vs 80-85 would probably be significant, not so much only on the cpu but on other system components as well. Not to mention the fan wear and tear and the fact that leakage of the transistors increase as temperature rises. And yeah, thermal paste is probably a significant cost factor. But on a computer with such as large pricetag you would expect something half decent, especially if >100 degree temperatures are seen under the stock thermal paste (and there are other threads on thermal repastes and heatsink quality showing a significant drop in temps).

Again, that was the heat sink. The heat sink is obviously not the CPU. No one in this thread has yet explained where the CPU core temperatures are coming from in some of these tests (whether that's Intel circuitry or something else), and until that's resolved, cross-manufacturer comparisons of temperature readings don't make much sense. The fact that I can't even get Mac programs that are supposed to read directly on the CPU itself to agree on temperatures (I saw a 15 degree band across two different programs) is, well, problematic. There's max and rolling averages, there's max for each core, etc. Apples to apples comparisons are important if we're going to make these kinds of arguments.

And per the above, your first sentence here remains a claim without a warrant. As for the 13" model, it's not a quad core. That in and of itself is super significant. Once again, valid comparisons are the hallmark of valid points.

Its intel's on die temperature sensor that gives the important reading. Not the heatsink (though there is a sensor for Tcase which is generally located near the heasink). I mentioned that before. Well of course the 13" is not a quad core, that wasn't the point. Well perhaps my first sentence needs a disclaimer; Going by industry standards, the rmbp does not posess adequate cooling or the bios is configured to run at non optimal temperatures [1].

[1] Non optimal being the generally accepted <90 degrees threshold that should be seen.

The fact the bill-p is hitting 105 under prime 95 is proof enough.

To be clear, I could not replicate notebookcheck's test. (see below for proof)

So either 1.2GHz is a glitch, or they (notebookcheck) have a bad lemon... like the OP does.

And I have proof below that temperature reported in a computer program is not reliable.

Because it's reporting 105C, and that would have caused the machine to shut down.

And they do. Either that or... I have a super Retina Macbook... that's far superior to everyone else's.

Yes.

Image

Please look at the above and tell me it doesn't have adequate cooling?

Just as an aside, the rMBP 13" does get as hot (CPU temperature reaching over 85C).

http://www.notebookcheck.net/Review-Apple-MacBook-Pro-Retina-13-Late-2013-Notebook.105035.0.html

What can I say? I know my machine is working as intended (see screenshot above), and that's more than enough for me.

Sensors can get weird readings sometimes. I have seen the CPU in my laptop supposedly draw 21 KW of power which is obviously wrong. That does not mean that the software utility is always wrong only that there may be bugs. It could be a decimal rounding problem. Note that its actually a difference tmax- temperature that must stay above 0 and there is also something called tj target which varies per CPU. Temp sensors may not be completely accurate to less than a degree.

It happens on other systems too. Just means something is wonky.

For instance.

 

[cirus]

Emphasis added. I'm starting to think this whole thing boils down to you and me saying, "It's adequate and working as intended," and cirus, the OP, and others essentially saying, "But it could be manufactured to run cooler/support more TurboBoosting under existing temperatures." If that's a correct assessment, then maybe there's room for agreement (other than the subjective claims as to whether there is, or is not, a "problem").

This is probably true. Mainly I believe that the high temperatures are not acceptable and should be lower. Hence something is off.

 

[iThinkergoiMac]

I don't have any rMBP with me to test right now (returned them), but I can create a sample project file for you to test if you have one available.

I appreciate the offer, but my MBP is from 2009, so that doesn't help much...

 

[maplingstorie]

**** gpus

 

[luffytubby]

It happens on other systems too. Just means something is wonky.

For instance.

Image

I was browsing the Intel support forum, and came across a thread from a dude with a HP laptop that had high temps. Here is what the intel rep responded to the guy and his high temperatures;

The TCase for this processor is 105 degrees Celsius.
The TCase is a number established by Intel® as a point of reference in order to understand what could be expected as per normal processor temperature.

Anything from the Tcase and below will be the expected temperature of the processor in normal use, anything that doesn’t stress out the processor (watching movies, burning CDs, browsing the internet, creating documents, etc.) When the processor is stressed out meaning that you are running heavy processor applications that take control of the CPU or uses it at 100% the temperature will go beyond the Tcase. It can perfectly reach 115 to 120 degrees and the processor will still be OK. The cooling fan is in charge to keep that temperature there.

If the processor temperature reaches 130 degrees or more it will send a signal to the motherboard to shut down to prevent mayor damages and most likely it won’t be possible to turn the computer back in until it cools down.

The normal processor temperature will depend on the chassis type, the hardware involved and the location of the computer, and it usually is lower than the Tcase.

So the temperatures that you are reporting are OK, also I see that you are reading the cores temperature, you actually need to check the CPU temperature, and not the cores temperature.
In order to do so, please check with the laptop manufacturer to see if they have thermal monitoring software that can monitor the CPU temperature instead of the cores temperature.

from this discussion; https://communities.intel.com/message/158824



This to me is interesting. Why is he saying that he should not read the core temperatures? Does anyone know? are they less of an indicator of the true heat, because there is a massive difference between those temps(100s) and those of normal CPU which is 78 (in your case).

 

[Gav Mack]

I think the MBP is pretty very high up there in terms of a notebook being a professional's computer, though the Retina models I feel falls badly short in terms of colour gamut despite the amazing resolution. My 17 inch 2010 model will eventually get replaced by a used last gen 17 inch with Thunderbolt as I prefer that display panel, along with being able to fit SSD upgrades etc.

I prefer to do the bulk of my work on my Mac Pro with a Dell matte LCD which says everything about which pro computer I like to work on the most

 

[hipnetic]

The OP quoted someone's post, but changed the headline. The person who wrote those comments used a title of "Hear me on this: Macbook Pro (Retina) is not for Photographers or Video Editors". The OP's decision to use the headline/title "What's your opinion on the MacBook Pro not being a professional's computer?" has gotten everyone riled up here about the definition of "professional."

 

[Swampus]

He points out valid criticisms and he's a troll?

I'm not a fan of the rMBP either, but, yeah, this is a troll. That's pretty much how the good ones operate. They include a lot of valid arguments and have a sort of "if only everyone was as smart as me attitude." That's how they generate a dozen pages of replies in only a few days.

This one is good, but not so good that s/he didn't get sockpuppets confused in the first hour:

 

[Swampus]

Emphasis added. I'm starting to think this whole thing boils down to you and me saying, "It's adequate and working as intended," and cirus, the OP, and others essentially saying, "But it could be manufactured to run cooler/support more TurboBoosting under existing temperatures." If that's a correct assessment, then maybe there's room for agreement (other than the subjective claims as to whether there is, or is not, a "problem").

For the discussion generated here, I think that really is a good summary. I'm a little old-school about temps, but that's often solved by simply redoing the thermal compound yourself. I find it annoying, but it's the same with any brand. If you want it to run cooler, you have to do it yourself.

 

[bill-p]

Sensors can get weird readings sometimes. I have seen the CPU in my laptop supposedly draw 21 KW of power which is obviously wrong. That does not mean that the software utility is always wrong only that there may be bugs. It could be a decimal rounding problem. Note that its actually a difference tmax- temperature that must stay above 0 and there is also something called tj target which varies per CPU. Temp sensors may not be completely accurate to less than a degree.

It happens on other systems too. Just means something is wonky.

Unless you say my sensors are buggy (in which case, that means I'm not getting 100C under load), I think it's clear that 100C still allows the rMBP to operate perfectly fine, and 105C won't cause the computer to shut down. In which case, core temperature reading is not reliable.

As an aside, someone just posted a link above as to why reading core temperature is not reliable, so I think that should make a good read.

Plus you just proved our (john123's and my) point with that screenshot there. See how different programs report different temperature numbers?

 

[john123]

I was browsing the Intel support forum, and came across a thread from a dude with a HP laptop that had high temps. Here is what the intel rep responded to the guy and his high temperatures;

from this discussion; https://communities.intel.com/message/158824

This to me is interesting. Why is he saying that he should not read the core temperatures? Does anyone know? are they less of an indicator of the true heat, because there is a massive difference between those temps(100s) and those of normal CPU which is 78 (in your case).

I think I can shed a little light here. Tcase is a spec generally provided to manufacturers so they can design their thermal profile. Tjunction is the value you'll see on the Intel's web site and is relevant because it's (roughly) the expected peak core temperature. I say "roughly" because Tjunction values reported are actually a target, rather than the (unreported) Tjunction Max (unreported). This is especially true in the multi-core context where there's some hidden balancing going on. Altogether, that's why the 105 degree reading posted by bill-p isn't really anomalous.

 

[john123]

The impact of 95-100 degrees for significant amounts of time (heavy rendering) vs 80-85 would probably be significant, not so much only on the cpu but on other system components as well. Not to mention the fan wear and tear and the fact that leakage of the transistors increase as temperature rises.

Since I think we concurred that there's common ground, I'm only going to respond to a small part of your post, because I think it sheds some light on the entire debate. I highlighted the important part. My gimmick experiment back in post #89 suggests that that impact actually isn't that big of a deal. At low CPU loads, my heat sink temperature is around 36 degrees. At 100% load with 8 threads (4 cores x 2 HT), I hit 46 degrees. So regardless of how hot the CPU and its cores themselves are, the rMBP is dissipating the vast majority of that heat.

The highest temperatures around that heat sink were obtained by kicking the 650M GPU up and (interestingly) not kicking the CPU up. (This was because the fans weren't going to their 6000 rpm max when I wasn't hitting the CPU, although perhaps the temperature readings suggest that they should). But here, proper application of Arctic Silver to the CPU won't actually make much difference. We could have a separate debate about cooling around the GPU, I suppose, but I'd rather not get off track.

Anyway, my point is that contrary to what you stated, the impact on other components doesn't seem likely to be large, even if the on-core temperature is staying around the Tjunction value. 46 degrees on the heat sink doesn't strike me as problematic. You might counter with, "But if they did it better, it might go down a few degrees, and those extra degrees might lead to X% more units failing," and I won't disagree with that. But then we're back in the semantic debate about what is and is not a "problem."

 

[zeego]

After taking years to finally switch over to Apple and specifically MacBooks, I have to admit I have not for one second looked back or otherwise thought of any Apple product not being in class on the market.

I spent thousands of dollars over years time on Windows laptops ranging from Toshiba, Dell, Samsung, Acer, etc. and most of which only lasted me upwards of a year before something died, failed or I was in need of an upgrade. Most the problems I experienced were due to heat either due to poor ventilation, cooling, or bad heat sync.

In fact one year I bought a brand new $1600 Toshiba "laptop" that I couldn't work more than 30 minutes with it in my lap before it would overheat and trigger the safety shutoff.

I was also one of those people that didn't want to get an iPhone right off because of the fanboy's and my father being an avid Apple fanatic (still has a working Apple II and Apple Lisa and every other Apple product under the son). Now I can honestly say I will never own a phone that isn't an iPhone. I loved being an Apple hater and claiming that if you wanted a real work machine it needed to be running Windows.

However, once I made that switch and started learning OS X in all it's glory as well as the quality and thorough thought process behind the design of every Apple product, I can now honestly admit I am one of those Apple fanboy's that I used to despise.

I don't think I'm in a higher class or anything because I use Apple products, and I don't look down on Windows or any of their machines. I still have a great custom Windows desktop computer that I have in my home office. It doesn't get near the use my Windows machines used to, in fact I mainly use it for the 2 27" monitors for the real estate and for playing the occasional game (even though they haven't released anything recently worth playing or paying for). I can admit though that my MBP gets most use and is my main go to machine for when I need to do most things that I consider to be "professional" work.

Long story short, I don't agree with the opinion of anyone saying that the MacBook Pro is not a professional's computer. In fact I would go as far as to say that I now look at the MBP as the official computer of most professionals.

 

[FrozenDarkness]

most laptops now have temperature sensors that auto-shuts off when it gets too hot. unless your worried your laptop is burning your skin there's nothing to worry about. i'll leave it up to engineers 10x smarter htan me

 

[phsphoenix]

I've looked over a few pages of this thread and seriously, MacSumo you're a troll.

For <insert PC here>, even the best ones, you will run into driver issues as soon as you try to install another OS or even reinstall the OS the computer came with, unless you want to use the built in recovery partition which typically puts all that OEM spyware back.

Obviously, you can build one yourself, but umm... good luck doing that with a laptop. If you do build a desktop yourself, you probably wouldn't be answering this thread.

I don't know how valuable your time is for you, but personally, the 10 hours that I have to spend hunting down each and every driver, especially for those OEMs who don't even offer them (HP I'm looking at you) could be spent doing something more productive, like coding.

That said, OS X does have it's flaws. It's a PITA to get a lot of FOSS software working on Mac OS X, and most of the time it's like Linux From Scratch, since you have to hack the configuration files yourself, but once you have something, it works, and continues working.

Also, the extra money that you spend on Apple hardware is worth it considering how many OS X versions that hardware will support. Apple ensures that the proper drivers ships with the OS for several years, and they don't suddenly decide to change the entire GUI (see Windows 8).

There is no computer that's 100% perfect, because perfection is in the eye of the beholder, and in Apple's case, the beholders consist of the entire planet. As far as coming close to perfection though, well, don't take my word for it, but Apple isn't one of the most profitable companies in the word for nothing, and no, it's not only because of their marketing department.