My mac book pro with retina just wouldn't turn on after barely 3 months after I bought it. Really bad experience. Lost all my information.
With regards to your display's power consumption, that 115W is not equivalent to a 100W incandescent light bulb. To begin with, fluorescent lamps are 5 - 6x more efficient than an incandescent lamp, and CCFL are about 50% better than that. That means, there are more lumens per watt emitted by the CCFL lamp; it can produce more light using much less wattage.
That 115W is distributed to power supplies for all components, including an inverter to provide the high-voltage the CCFL tube needs. There is definitely power lost to heat in power supplies, though I don't have an efficiency number handy. The LCD and its drive electronics are much more efficient, i.e. they dissipate less heat per watt [than a power supply would].
To get an idea of the wattage which is dissipated as heat, the easiest way would be to measure it. I'm certain you would find the temperatures much, much, lower than a 100W incandescent lamp (which, creates its light by heating a filament - so most of its power consumption is by definition converted to heat). If your display were to be as inefficient, there wouldn't be enough power left for the display components - it would find better work as a toaster
I am well aware of most of what you posted regarding LCD theory and tech. In my previous life I was VP of sales for an OEM rep firm who sold LCD displays from Sharp, Sumitomo, GE and LG Philips to companies including HP, Microsoft, Boeing, Seimens, Fluke,Tektronics, Nintendo, Adobe, intel etc. As for the relevant IR issue here I'm done getting into this tech debate. Bottom line is my business owns 6 MBPr (4 LG, 2 Samy) and there are no signs of IR on any of them (so far). I am monitoring this thread to keep tabs on any updates / changes etc. I have no interest in continuing to therorize as to the cause of the ongoing LG IR problem, even if the cause is figured out there is almost zero chance it will be anything a user can do to 'fix' it on their own.
You either live in the Arctic or have the coolest running Quad core CPU in existence, lol. I have the same CPU as you (2.7Ghz 8MB Cache on all our MBPr's) and idle temps average 56C, this one is currently 58C with just safari open and 81F ambient (fans at idle 2160 / 2000). At full load of near 100% on all 8 virtual cores my CPU core temps will fluctuate from 90-104C (measured using Temperature Guage App). I'm not sure what you used for the CPU load during your measurments but the core temps you posted aren't possible during true 100% load. If you want an easy way to get 100% load on all 8 threads just run 8 games of Chess vs. Computer (defualt settings) at the same time with no other apps running. You can check your CPU load for all 8 threads by opening Activity Monitor and then hit 'command' + 2 (or window>CPU usage), make sure to minimize Activity monitor or it will reserve some CPU cycles for monitoring, as will any other open apps or processes including wifi, bluetooth etc. I bet you'll see a steep increase in max temps trying this. EDIT: note that if your using VM software the CPU will be throttled and not acheive true 100% loads (I think turbo boost gets disabled).
As to your comments about the heat of the CCFL RGB displays, your overthinking it. Basic physics dictate that a watt is a watt, it doesn't matter how it's generated or consumed, if the display is using 115 watts then it must dissipate 115 watts of heat. Since all the energy consuming components are within 2" of the back of the actual LCD display it doesn't really matter what is using the energy or how efficient it is. The 115 watts in the given enclosed area will generate the same overall temps in the space whether it's a 115W incandescent bulb, a 115 watt toaster oven element or the actual component parts consuming 115 watts. Yes, the first 2 would have hot spots but the overall temps inside the same given space will be the same. BTW, yes, the CCFL backlights are still tubes, I think 5 parallel tubes on the NEC and there is a difuser panel between the tubes and display. Also, FWIW, the front LCD panel and the rear cover appear to be the same temps after a full day of use, again a watt is a watt.
Sorry you feel drawn into the discussion as to the causes of IR. I'm not even asking that question here. I only was positing a question to possibly get an idea if there were any connection between extreme temperature and delayed IR. Since you brought up the topic of background, I'm a retired engineer with some 30 years experience in both electronic design and software engineering; also, I hold an Amateur Extra Amateur Radio license (and a few other less related licenses and certs).
My rMBP also has a perfect display. (Not much if a paragraph, but thought it worth mentioning...)
Yes, my ambient room temperature is cool, though not quite arctic, at a more modest 67° F. All 8 cores were being utilized at above 95% (and that's a snapshot taken during the context switch to the Hardware Monitor app). Activity Monitor graphs show all 8 at ~100%, though as you know, there isn't much resolution in those. I could post screen shots, but you already have the data. Think about the fan speeds also - they would not be running at maximum RPM if there weren't a high thermal load to dissipate. The application I used to load up the CPUs is a popular video conversion tool (it's not necessarily appropriate to discuss some of those apps here).
I'm definitely in the category of "power user", so I'm well aware of how to use Activity Monitor, or a variety of command-line tools - they would all show approximately the same loads. I could also have started multiple instances of 'yes' - that probably would have increased the load slightly, but it doesn't really matter. I did the measurement to see how much heat might be found near the display, as you suggested. A slightly higher load wouldn't raise the external temps by much.
I'll still have to disagree with the statement "a watt is a watt". It is not strictly a measurement of heat dissipation. If you have an amplifier that can drive a speaker with 20 watts of power, but uses 40 watts to accomplish this, the difference is lost as heat - but the 20 watts used to drive the speaker goes into mechanical work (and some heat loss in the speaker coil). Anyway, my point is the wattage consumed by a device is a combination of work done and heat loss - the goal being to get the most work done. Another example is shrinking semiconductor die processes - going from Sandy Bridge to Ivy Bridge does the same work, but the latter consumes less power because of lower resistance (because of smaller features), which directly equates to less heat dissipated. Conversely, it could do more work at the same power consumption (all other changes between the chipsets ignored).
It's been a fun discussion, but it's probably more appropriate to let the thread get back on topic.
That said, I'm still interested in any feedback from those with IR that occurred after a long period of use and who might remember if their machine had been exposed very high temperatures.
Yes, that's an interesting suggestion.
One thing that came to mind that might have skewed my reported temps (compared with what you see) is my configuration of Hardware Monitor isn't looking at the CPU temp diode - rather, it's reporting the CPU Proximity sensor, which is undoubtedly much lower than the internal device sensor.
I ran the suggested 8 games of Chess; computer v computer, 256 moves ahead (just for extra measure?).
Verified 100% core activity for all 8.
Fans at maximum.
Ambient temp 66º F / 18.8º C
Nvidia Discrete Only (Auto GPU switching disabled)
CPU A Proximity
153º F / 67.2º C
151º F / 66.1º C
GPU Temp Diode
172º F / 77.8º C
165º F / 73.9º C
main heatsink 2
120º F / 48.9º C
120º F / 48.9º C
main heatsink 3
131º F / 55º C
127º F / 52.8º C
Bottom line - the temps didn't increase all that much. And with regards to our discussion of display temps, the heatsinks (and subsequently the exhaust temps) only increased by a few degrees.
(Apologies to the OP for being slightly off-topic again. We now return you to the thread already in progress.)
I just bought a 13'3 MBPr this morning in the apple store :13'3/2.6/256/8, it's an early 2013 one.
The command line to see the screen's manufacturer doesn't work, but I've made the ten minutes test with the black and white boxes and switching to the grey font, and no IR at all. Very happy, it's an awesome and wonderful tool !
I hope it wont. I bought the old version at a high tech store sunday, and i got my money back this morning, because it had IR right out the box. It was the 256 go 13'3, but the old version ( before 2013).
That's why I went to the apple store, hoping to get one without IR. And this one, after 2/3 tests of 45 min of black and white squares is still perfect.
So it may pop later, but for now, it's perfect to me
After near 2 months of use, here are the news :
I'm using the Mbr everyday since I bought it, for various use : internet surf, some little games, and a lot of typing (i'm studying and working my law degree with it).
I just redid the checkerboard test, 10 min, and still NO image retention at all. So after reading all of you, I guess I'm lucky.
I dunno if it is a samsung or LG one, but i love the screen and haven't seen anything wrong since I got it.
No IR, no light problem, no shadow... crossing fingers.
I'll redo a test in a few months.
Ooops, my mistake I misread the CPU "proximity" sensor line in your original post. Not sure how long you ran the high load test before measuring temps. My long periods of high load usage (2-3 hrs continuous) may also increase the surface temps of the display to a higher level than you measured, I never really considered how long it might take for the vent output to maximize the surface temps of the display. Also my vent output definately blows more exhaust air onto the front of the dipslay than under the hinge, I even removed my my speck satin case to make sure it wasn't redirecting the airflow. That's with the display open to approx 20 degree past vertical.
I will say that your Ivybridge vs. sandybridge is a perfect example of efficiency but not energy dissipation. The new smaller arcitecture does do the same amount of work in a smaller package, while using fewer watts to due it. A watt is a measure of energy expressed by temperature (heat) 1 Wh = 3.412 BTU's it's a simple linear equation. No way around the fact that if a device consumes 100 Watts of energy it must dissipate 100 Watts of energy to return to the same temperature it started at. A Watt is a Watt and what goes in must come out to return to the same state, doesn't matter if the output is electrical, mechanical or light, the energy never dissapears it is just transferred.
I only ran the test until the temperatures more or less stabilized. I monitored them until the curve showed a cooling trend, then flattened out. Looking at the graph, I'm seeing all this happen within about 6 minutes.
I can't remember what angle my display was at when I was checking the flow. I just watched the movement of the down and it seemed to indicate the flow I described. Anyway, it's kinda moot, as the air just isn't all that hot.
A Watt is a Watt and what goes in must come out to return to the same state, doesn't matter if the output is electrical, mechanical or light, the energy never dissapears it is just transferred.
Yes, that's what I was saying . Energy is conserved, but can be converted to different forms. A watt can be used to refer to heat (light), but is not necessarily so.
May all our displays remain IR free.
Yes but a W.h. still = 3.412 BTU's of heat in any of those forms of energy. No matter which form it's in the heat produced is the same amount per W.h. consumed. So in the case of our CCFL backlit display the 'typical' 115W power draw (or whatever the actual draw is) is the same amount of heat that has to be dissipated. This is especially true with light, any 115W light, consumes and produces exactly 115W of heat (minus wire resistance etc.), if it created any more than 115W of heat then you would have discovered perpetual energy ..... It's the same with a 115W toaster element, 115W speaker (hard to draw a constant wattage) or a 115W CPU. A Watt is always a Watt and the heat per watt is constant.
PS: To be technically precise all the above 'W's should be expressed in Wh (Watt Hours) since time is required to measure transfered energy.
Hello all, recap and conclusion.
Got my Mac Book Pro Retina in the first week of release. LG
Discovered IR after three months.
Had screen replaced with no trouble, LG
Discovered IR AGAIN after three months.
Asked to have the screen displace replaced.
I was told it was normal and it would not be replaced.
I talked to three Apple Care people everyone higher in position than the next.
They all said no replacement and I could not speak to anyone else.
I sent screen shots to Apple of this thread with people having their screens replace.
I received a call from Apple telling me they would replace the screen.
Took it to the reseller and they replaced it.
I did a jig (dance) in the store when I ran the terminal command and found a Samsung.
Checked for dead pixels and yellow tinges. All seems good.
Checked Apple logo on back of screen, it's a little pink/purple but I can live with it.
I am a graphic designer/photographer so there was no way I could live with IR
IR is not normal don't let them tell you other wise.
Good luck to all that have to go through this nightmare.
P.S. What a stressful and waste of two weeks with this issue.
I still love Apple but will think twice before I buy 1st generation hardware.