Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Audio Decoding KILLS MacPro (2009) green factor
- Thread starter handheldgames
- Start date
- Sort by reaction score
Some More Info
I saw this on
http://episteme.arstechnica.com/eve/forums/a/tpc/f/8300945231/m/636006172041
Originally posted by mckill:
if it is the power management, you could simply try disabling it temporarily by remove AppleIntelCPUPowerManagement.kext in /System/Library/Extensions/ and rebooting.
Hi Mckill
I have been watching problem since seeing a post on Sound on Sound about Dec 15. and also Apple Support Discussions .
I Have a Macpro 2009 Dual 2.26 and have been able to to get immediate Temp rise simply
by playing ITunes for as little as 5 mins . Also by just turning on my Presonus Firewire
Audio interface and doing nothing .
Until you posted this suggestion no one had seemed to know what to do .
I have two partitions one with Leopard 10.58 and one with Snow Leopard 10.62 and both
do exactly the same .
So i followed your suggestion and removed (AppleIntelCPUPowerManagement.kext) From my Snow Leopard and rebooted .
Then turned on Firewire Audio interface and watched Temps for 15 mins .
The result was NO Temperature rise at all , Then also ran Itunes with it for 15 mins
The result was CpuA 30C to 32C .CpuB 27C to 28C .
I then replaced (AppleIntelCPUPowerManagement.kext) and all problems of temp rises of over 41c were back again .
But it is Defiantly what is Causing the Problem .
Removing this Kext also creates other Problems -Because it controls all Powermangement - such as no Sleep for the computer . But as i use the mac for Audio mainly i never put it to sleep .
I hope this might help with pointing toward someone who can Program A new Kext that will work .
Contacted Apple Europe Support and passed on my Findings . ?? - .
Doing some more testing in a few days and will post up Istatmenus figures .
Laters
I saw this on
http://episteme.arstechnica.com/eve/forums/a/tpc/f/8300945231/m/636006172041
Originally posted by mckill:
if it is the power management, you could simply try disabling it temporarily by remove AppleIntelCPUPowerManagement.kext in /System/Library/Extensions/ and rebooting.
Hi Mckill
I have been watching problem since seeing a post on Sound on Sound about Dec 15. and also Apple Support Discussions .
I Have a Macpro 2009 Dual 2.26 and have been able to to get immediate Temp rise simply
by playing ITunes for as little as 5 mins . Also by just turning on my Presonus Firewire
Audio interface and doing nothing .
Until you posted this suggestion no one had seemed to know what to do .
I have two partitions one with Leopard 10.58 and one with Snow Leopard 10.62 and both
do exactly the same .
So i followed your suggestion and removed (AppleIntelCPUPowerManagement.kext) From my Snow Leopard and rebooted .
Then turned on Firewire Audio interface and watched Temps for 15 mins .
The result was NO Temperature rise at all , Then also ran Itunes with it for 15 mins
The result was CpuA 30C to 32C .CpuB 27C to 28C .
I then replaced (AppleIntelCPUPowerManagement.kext) and all problems of temp rises of over 41c were back again .
But it is Defiantly what is Causing the Problem .
Removing this Kext also creates other Problems -Because it controls all Powermangement - such as no Sleep for the computer . But as i use the mac for Audio mainly i never put it to sleep .
I hope this might help with pointing toward someone who can Program A new Kext that will work .
Contacted Apple Europe Support and passed on my Findings . ?? - .
Doing some more testing in a few days and will post up Istatmenus figures .
Laters
im a young "in" person.cl if you need any help modifying the looks of the page i might be of some help.
i think there needs to be some sort of introduction on the front page. something friendly and straight to the point. facebook is not meant to be a novel.
to be perfectly honest, having a title of "my mac pro '09 heats up when play audio" and leaving it at that would be perfectly suitable for facebook. but you know...
Crikey, so compressing the mission statement into 31 words leaves it still too verbose! Eyes rolling here too - as in, I so do not know how to move the Facebook levers.
Perhaps DoFoT9, you could pick up the ball ...
You are welcome to all the material, please take the project on if you think you have the ability to make and sustain it. I'll leave my material a while and trash it if there's no take up/contributions.
How about this
Title:
Audio and Firewire turns my Mac Pro 2009 into a Heater and reduces Performance signficantly
Symptoms:
- Decrease in performance (up to 25%) when rendering or other high performance computing
- Increase in CPU temps (beyond what should happen for 0.1% CPU utilization)
Performance Test:
1) Download Geekbench
2) Close all applications and run it, note your score
3) Play 3 songs in iTunes and run Geekbench again while they are playing
4) Note the score is reduced significantly for only 0.1% CPU utilization by iTunes
Temp Test:
1) Download Temperature Monitor / iStat
2) Check temperature of CPUs
3) Play 3 songs in iTunes (~15 min)
4) Check temperature of CPUs again
Title:
Audio and Firewire turns my Mac Pro 2009 into a Heater and reduces Performance signficantly
Symptoms:
- Decrease in performance (up to 25%) when rendering or other high performance computing
- Increase in CPU temps (beyond what should happen for 0.1% CPU utilization)
Performance Test:
1) Download Geekbench
2) Close all applications and run it, note your score
3) Play 3 songs in iTunes and run Geekbench again while they are playing
4) Note the score is reduced significantly for only 0.1% CPU utilization by iTunes
Temp Test:
1) Download Temperature Monitor / iStat
2) Check temperature of CPUs
3) Play 3 songs in iTunes (~15 min)
4) Check temperature of CPUs again
Good stuff. It feels as if this is progress.
>Here I am copying 172Gb of video files to one of the internal HDs...
>http://www.thunder-keep.co.uk/site/m...800hd_test.jpg
>...
>http://www.thunder-keep.co.uk/site/m..._tests_1&2.jpg
>...
>http://www.thunder-keep.co.uk/site/m...mperatures.jpg
URLs not working for me.
Maybe not so relevant, a video of iTunes plus work load:
http://www.youtube.com/watch?v=QwtaeFJ9MwA
(anyone know how to write a speeded up video?)
We need those refs to Intel's CPU spec that details temperature safety ceiling. (Just spent an age hunting for my own post detailing the same, and am drowning in volume here and on apple discussion list).
Also (per Concorde's insight), suggest we prioritorize for performance hit before other symptoms (temp, power draw, etc).
Also (per Concorde's insight), suggest we prioritorize for performance hit before other symptoms (temp, power draw, etc).
Let's stress the performance degradation
Hello, brave also-Apple-customers, please excuse if my English isn't 100% polished since it's late in the day and I'm not a native speaker.
At first, big compliment for your great work so far. Work that had been Apple's job! But they also seem to have to save money in those days so let us do the beta-testing. A strategy which was traditionally attributed to Microsoft and one of the great arguments that Apple thinks different (I hoped so).
I've read this whole thread (which took some days ) since I also have temperature issues on my MacBook (fresh end 2009, 2.26 GHz C2D, polycarbonate unibody, OS X SL, my first Mac!), however not with respect to audio. So my issue only seems to be the awkward fan management in OS X SL (one of your 4-5 points, see below).
As far as I've read this is a problem since years with Apple's notebooks I am surprised that your MPs are also affected as they're a) the premium and b) desktop systems of Apple where the cases should be spaceous enough to allow good cooling. I always heard about the great cooling concept in the Mac Pros (6 fans, each one big enough that it can run with less rpm and this way produce less noise).
When doing yes in the terminal, for one time I drove the CPU diode a few deg over 100°C (iStat Pro, Hardwaremonitor) at which point I of course stopped, fearing damage on my system. - The fan did nothing, constantly remaining at 2000 rpm. Being a friend of a low noise system, my impression is that Apple here sacrifices safety for this design goal (maybe intentionally braking the fan triggering in OS X?). I didn't do heavy work with it yet, but wanted to use it for software development in the near future. And I fear heavy CPU load e.g. while compiling could have the same effect.
I am angry like you that they are so little customer oriented in this issue (well, still the hope: the lady of the European support who Dragonforce contacted). For the money I paid and the image of premium quality out of one hand (hw+sw) Apple cultivates I can expect a reasonable ventilation concept even in my notebook. Well, as you already wrote.
I find it bold to drive the CPUs at so much heat (maybe intending that they fail earlier so that people buy their next Mac earlier...?) and tell us it's within limits.
@ByteOfApple: Your headline already sounds good (especially the turns my Mac into a heater) I think - as smacman pointed out so often and cl-user just restated above our best argument is the performance loss so we should put that first. Because else it may reads as being too harmless (someone may read the headline and think... well, okay, some audio issues... I'm not concerned). To prevent this and make clear that potentially anyone is affected by this bug in OS X (since under Windows this issue doesn't exist) wasting 50 W and making Apple non-green - I propose to put the headline in a more severe tone. My suggestion (by looking at smogsys bug report):
25% Performance Degradation Due to Excessive Power Consumption in 2009s Mac Pro
So when this will appear on the NYT headline, this should work. Well, just my thoughts.
@smogsy: Already great, too, I think. And also here I would put performance loss first - as the heaviest argument (for the business people at Apple we have to wake up somehow and since they deny the temperature aspect of our problem). So the performance degradation you mentioned in the description in my opinion should also appear straight in the title, maybe the headline I proposed above could serve our purposes here, too. And for the problem description (just little resorted and paraphrased):
>Description of the Problem:
2009 Mac Pro family computers suffer performance degradation up to 25%, consume excessive power and thus generate excessive heat during any simple audio playback and/or some firewire traffic, although being at low CPU load.
1.25% performance degradation (benchmarks)
2.Low CPU load (as expected on today's machines) during simple audio output of any kind (not only iTunes) as indicated in Activity Monitor.
3.Excessive, near 100% power consumption of CPU (~50W) ← Apple's non-green!
4.Heat generation up to near 100% load temps (70-80C)
5.No fan speed increase
I hope my suggestions are helpful. I hope they would not only fix the strange power consumption bug but also the fan management in OS X (e.g. making it configurable out of the box instead of our having to download and install third party software like Fan Control).
P.S.: Unfortunate that the AppleIntelCPUPowerManagement.kext is not configurable.
Hello, brave also-Apple-customers, please excuse if my English isn't 100% polished since it's late in the day and I'm not a native speaker.
At first, big compliment for your great work so far. Work that had been Apple's job! But they also seem to have to save money in those days so let us do the beta-testing. A strategy which was traditionally attributed to Microsoft and one of the great arguments that Apple thinks different (I hoped so).
I've read this whole thread (which took some days
) since I also have temperature issues on my MacBook (fresh end 2009, 2.26 GHz C2D, polycarbonate unibody, OS X SL, my first Mac!), however not with respect to audio. So my issue only seems to be the awkward fan management in OS X SL (one of your 4-5 points, see below).As far as I've read this is a problem since years with Apple's notebooks I am surprised that your MPs are also affected as they're a) the premium and b) desktop systems of Apple where the cases should be spaceous enough to allow good cooling. I always heard about the great cooling concept in the Mac Pros (6 fans, each one big enough that it can run with less rpm and this way produce less noise).
When doing yes in the terminal, for one time I drove the CPU diode a few deg over 100°C (iStat Pro, Hardwaremonitor) at which point I of course stopped, fearing damage on my system. - The fan did nothing, constantly remaining at 2000 rpm. Being a friend of a low noise system, my impression is that Apple here sacrifices safety for this design goal (maybe intentionally braking the fan triggering in OS X?). I didn't do heavy work with it yet, but wanted to use it for software development in the near future. And I fear heavy CPU load e.g. while compiling could have the same effect.
I am angry like you that they are so little customer oriented in this issue (well, still the hope: the lady of the European support who Dragonforce contacted). For the money I paid and the image of premium quality out of one hand (hw+sw) Apple cultivates I can expect a reasonable ventilation concept even in my notebook. Well, as you already wrote.
I find it bold to drive the CPUs at so much heat (maybe intending that they fail earlier so that people buy their next Mac earlier...?) and tell us it's within limits.
@ByteOfApple: Your headline already sounds good (especially the turns my Mac into a heater) I think - as smacman pointed out so often and cl-user just restated above our best argument is the performance loss so we should put that first. Because else it may reads as being too harmless (someone may read the headline and think... well, okay, some audio issues... I'm not concerned). To prevent this and make clear that potentially anyone is affected by this bug in OS X (since under Windows this issue doesn't exist) wasting 50 W and making Apple non-green - I propose to put the headline in a more severe tone. My suggestion (by looking at smogsys bug report):
25% Performance Degradation Due to Excessive Power Consumption in 2009s Mac Pro
So when this will appear on the NYT headline
, this should work. Well, just my thoughts.@smogsy: Already great, too, I think. And also here I would put performance loss first - as the heaviest argument (for the business people at Apple we have to wake up somehow and since they deny the temperature aspect of our problem). So the performance degradation you mentioned in the description in my opinion should also appear straight in the title, maybe the headline I proposed above could serve our purposes here, too. And for the problem description (just little resorted and paraphrased):
>Description of the Problem:
2009 Mac Pro family computers suffer performance degradation up to 25%, consume excessive power and thus generate excessive heat during any simple audio playback and/or some firewire traffic, although being at low CPU load.
1.25% performance degradation (benchmarks)
2.Low CPU load (as expected on today's machines) during simple audio output of any kind (not only iTunes) as indicated in Activity Monitor.
3.Excessive, near 100% power consumption of CPU (~50W) ← Apple's non-green!
4.Heat generation up to near 100% load temps (70-80C)
5.No fan speed increase
I hope my suggestions are helpful. I hope they would not only fix the strange power consumption bug but also the fan management in OS X (e.g. making it configurable out of the box instead of our having to download and install third party software like Fan Control).
P.S.: Unfortunate that the AppleIntelCPUPowerManagement.kext is not configurable.
the mission statement is fine. but the title of the page needs to be precise, to the point, and in facebook terms - much like what i suggested before
i think i can help, as long as nobody has any objections that is. i do not have the technical no-how to write up a decent explanation, but i should be able to format it so that it speaks to most people on the internet, old and young alike!
send me over an email cl (check my profile) with anything further that you would like to chat about, or put it in here. i dont mind
looks great!
for this facebook group, i think the main "focus" of it should be to get as many users as possible to join - from there we can show the number of users joined( and any comments they have made) directly to apple, this might "wake them up".
it would also be a good idea to have regular (weekly?) "status updates" to inform group members what is happening, or of any updates that there may be.
I suggest DoFoT9 that you just make a Facebook space for the topic. Please grab any/all material that suites your perspective.
My sense is that the average 09 Mac Pro user is either a professional or a very serious amateur user. Hence the audience concerned with this issue is likely to be quite mature and relatively wealthy, hence kids' speak might not lure in many.
Given that the typical response by users who are unfamiliar with the issue is denial, winning words could be a challenge.
I get the impression that we have two messages for 2 different audiences. One to get through to Apple, the other to win the participation of users.
My sense is that a professionally styled web site which includes an easy way for people to participate/contribute (at least sign up in some way) might suit the latter purpose. Whether a Facebook presence can suffice I don't know.
I also think it would be really useful if such a site could act to channel the anger that Apple generates by operating in such a way that we users feel snubbed and frustrated. Can we redirect that user energy into some form of petition effort, a stronger set of knuckles with which to knock on Apple's door, wake them up. At least coax or trick Apple into letting us know what is going on (if anything).
Our simple end goal - get a MacOS bug fix asap.
for this reason i think that the facebook group should not be the main site that users go to. i doubt that most pro users would have a facebook account.
nope i agree with you. perhaps we should follow deadtimecapsule.org?
it would be great to have a form that users could fill out, much like DTC.org, and the results given back to the users. we could then also have a facebook group where our more colloquial users are. maybe?
as for the facebook group im in just tell me then group name once created
As above:
http://www.facebook.com/pages/2009-Mac-Pro-audiofirewirepower-bug/295274123618
Agreed.
Excessive Power Usage and Temperature rise are SYMPTOMS *NOT* the *CAUSE*.
If images are needed PM me.
I'm currently working with my Executive Relations person on this issue!
A bit naughty, but perhaps Apple are busy elsewhere today ...
http://www.flickr.com/photos/45223615@N05/4309409354/
THIS PICTURE IS BRILLIANT !!!!!!!!
You're too kind. I tweeked it, so new URL
http://www.flickr.com/photos/45223615@N05/4309409354/
(hope not a step too far)
Okay this was the first time running Cinebench so I hope I did this correctly; I just ran all tests.
Specs are: 2009 Mac Pro 8x2.66GHz, 12GB RAM, 2x1 TB HD, ATi Radeon 4870
1. No Audio/HDs
OpenGL: 5271
Single CPU Render: 2832
Multiple CPU Render Test: 21910
2. Audio(iTunes playing an .mp3 loop)
Open GL: 5271
Single CPU Render: 2819
Multiple CPU Render Test: 21060
3. FW800 Transfer(WD Studio Edition)/No Audio
OpenGL: 5087
Single CPU Render: 2817
Multiple CPU Render Test: 21913
4. No Audio/HDs
OpenGL: 5370
Single CPU Render: 2821
Multiple CPU Render: 22380
I forgot to log the starting temp, but it generally is around the low 30 C without audio but the machine averaged 61 C, with a high of 63 C when the music and hard drive portions of the test were conducted.
Specs are: 2009 Mac Pro 8x2.66GHz, 12GB RAM, 2x1 TB HD, ATi Radeon 4870
1. No Audio/HDs
OpenGL: 5271
Single CPU Render: 2832
Multiple CPU Render Test: 21910
2. Audio(iTunes playing an .mp3 loop)
Open GL: 5271
Single CPU Render: 2819
Multiple CPU Render Test: 21060
3. FW800 Transfer(WD Studio Edition)/No Audio
OpenGL: 5087
Single CPU Render: 2817
Multiple CPU Render Test: 21913
4. No Audio/HDs
OpenGL: 5370
Single CPU Render: 2821
Multiple CPU Render: 22380
I forgot to log the starting temp, but it generally is around the low 30 C without audio but the machine averaged 61 C, with a high of 63 C when the music and hard drive portions of the test were conducted.