I suggest the easiest and probably fastest way is to play in the last backup of your system drive, which you hopefully created just before you played in Cindoris patch!
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
I suggest the easiest and probably fastest way is to play in the last backup of your system drive, which you hopefully created just before you played in Cindoris patch!
Oh great. Another one in denial, who jumps into this thread without reading the previous posts. You have conducted one test to check for a performance loss. Had you bothered to read the previous posts in this thread, you would have noted that several users including myself have dedicated lots of time to this, and confirmed a performance loss in multiple tests. You also would have noted that the temperature rise was confirmed with an IR thermometer. A power meter also confirmed the 1600% increase in CPU wattage when playing an audio file.
If you are so happy with your brand new quad 3.33, then get lost. There is a real issue here, and the rest of us want it fixed.
Hi Smacman,
You know who here... Is that why I haven't seen you at the ADP site lately?
Did you catch my little aside to the Times?
Hi! Been on the road for a while. I'm just bringing myself back up to speed now. What a depressing hobby this has become...
23.4W for playing a single mp3 seems to be a little higher for me. Am i wrong?
I'm not saying it's not a good work. It's a hell of a good work. I wish I was as talented a cindori. He is on the right track. My only concern is that this beta patch would be only hiding the problem instead of solving it... idk... I mean.. it is a beta no?
Smacman, welcome back! Be sure to talk to Chris from Ars Technica if you haven't already!!
Hi Smacman!
You're so right about this hobby of ours (and boy, a lot others). I keep promising to myself and other posters I won't continue to post and then I go and do it again. It was good to see you again though and I thought I should tell you and the others here about my latest experience with Apple which happened just today. No big deal, but interesting I thought.
As you know, I opened a case with Apple too. Like many, after being promised they'd send my info to the engineers and get back to me, they didn't. I called them today with my case number and asked what was going on with my case. As usual, they start from scratch all over again despite whatever the previous tech wrote down or not. So being pretty well up on things now and not in denial anymore as one poster here says, I shot down his explanations pretty darn quickly and left him little room to maneuver.
I asked what the engineers had said only to find that the other tech lied and never sent anything but an FYI to them. From the tech I talked to today what the other should have sent was I think he said was an RPA. Not sure what the initials stand for but it's asking the engineers for a definite reply back to them, to the tech who sends one to them.
To do that (and this is what I found interesting because I wonder if this was given to you or others) he sent me an Application to run when I could reproduce the problems. The App is called, "Capture Data".
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at125947AM.png
Unless I missed it, and that could be easy to do, I hadn't heard of anyone being sent this app before in regards to all this.
Well, long story short, it was easy to just launch iTunes, play a song, and see my temps jump 20 degrees. When they seemed to hit their peak I started the Capture Data" app which just took a minute or so to run. I ended up with a dmg. to send him and when I opened it up myself I found this folder. Pretty impressive I thought.
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at14100AM.png
I would think there's enough data there for them to work with, if they decide to.
So I just wanted to tell you and the others here this little story as part of the ongoing saga, knowing full well that where it will lead is probably nowhere. But I'll post back if they do, so we'll see.
Hope your doing well, and that you've gotten some rest. And yes, I believe you are the ideal person to contact Ars Techica if that was legit. Getting them to report on this would be a real coup.
Also, I keep seeing someone say the maximum temp is 70C for our machines. It's not, both Apple and Hardware Monitor say it's 95C. I don't want to argue that though, I just want people whose temps hit like 65-70C to know they can relax a bit if temps are their main concern. It's still a concern and so is the rest that has been written about all of this, but for me and a few others whose core temps can sometimes hit 90C , our concern there is pretty critical.
Take care, bud
Steve, Sam, whoever...
You're so right about this hobby of ours (and boy, a lot others). I keep promising to myself and other posters I won't continue to post and then I go and do it again. It was good to see you again though and I thought I should tell you and the others here about my latest experience with Apple which happened just today. No big deal, but interesting I thought.
As you know, I opened a case with Apple too. Like many, after being promised they'd send my info to the engineers and get back to me, they didn't. I called them today with my case number and asked what was going on with my case. As usual, they start from scratch all over again despite whatever the previous tech wrote down or not. So being pretty well up on things now and not in denial anymore as one poster here says, I shot down his explanations pretty darn quickly and left him little room to maneuver.
I asked what the engineers had said only to find that the other tech lied and never sent anything but an FYI to them. From the tech I talked to today what the other should have sent was I think he said was an RPA. Not sure what the initials stand for but it's asking the engineers for a definite reply back to them, to the tech who sends one to them.
To do that (and this is what I found interesting because I wonder if this was given to you or others) he sent me an Application to run when I could reproduce the problems. The App is called, "Capture Data".
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at125947AM.png
Unless I missed it, and that could be easy to do, I hadn't heard of anyone being sent this app before in regards to all this.
Well, long story short, it was easy to just launch iTunes, play a song, and see my temps jump 20 degrees. When they seemed to hit their peak I started the Capture Data" app which just took a minute or so to run. I ended up with a dmg. to send him and when I opened it up myself I found this folder. Pretty impressive I thought.
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at14100AM.png
I would think there's enough data there for them to work with, if they decide to.
So I just wanted to tell you and the others here this little story as part of the ongoing saga, knowing full well that where it will lead is probably nowhere. But I'll post back if they do, so we'll see.
Hope your doing well, and that you've gotten some rest. And yes, I believe you are the ideal person to contact Ars Techica if that was legit. Getting them to report on this would be a real coup.
Also, I keep seeing someone say the maximum temp is 70C for our machines. It's not, both Apple and Hardware Monitor say it's 95C. I don't want to argue that though, I just want people whose temps hit like 65-70C to know they can relax a bit if temps are their main concern. It's still a concern and so is the rest that has been written about all of this, but for me and a few others whose core temps can sometimes hit 90C , our concern there is pretty critical.
Take care, bud
Steve, Sam, whoever...
Hey, I installed OS 10.6 again and I still got there error. It is somehow different... Please somebody help, I don't want to format the disc if possible.
looks like it has something to do with sx.triaxis.kext.SleepEnabler(1.0.1)@0x6c535000->0x6c536fff
Attachments
Thank you Cindori, I remember I checked 10.6.2 as it is the newest version I have installed. I'll install OS X on an external Harddrive and then remove the Sleepenabler.kext file. Do I have to replace it with a new one or will it just work ?
You're so right about this hobby of ours (and boy, a lot others). I keep promising to myself and other posters I won't continue to post and then I go and do it again. It was good to see you again though and I thought I should tell you and the others here about my latest experience with Apple which happened just today. No big deal, but interesting I thought.
As you know, I opened a case with Apple too. Like many, after being promised they'd send my info to the engineers and get back to me, they didn't. I called them today with my case number and asked what was going on with my case. As usual, they start from scratch all over again despite whatever the previous tech wrote down or not. So being pretty well up on things now and not in denial anymore as one poster here says, I shot down his explanations pretty darn quickly and left him little room to maneuver.
I asked what the engineers had said only to find that the other tech lied and never sent anything but an FYI to them. From the tech I talked to today what the other should have sent was I think he said was an RPA. Not sure what the initials stand for but it's asking the engineers for a definite reply back to them, to the tech who sends one to them.
To do that (and this is what I found interesting because I wonder if this was given to you or others) he sent me an Application to run when I could reproduce the problems. The App is called, "Capture Data".
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at125947AM.png
Unless I missed it, and that could be easy to do, I hadn't heard of anyone being sent this app before in regards to all this.
Well, long story short, it was easy to just launch iTunes, play a song, and see my temps jump 20 degrees. When they seemed to hit their peak I started the Capture Data" app which just took a minute or so to run. I ended up with a dmg. to send him and when I opened it up myself I found this folder. Pretty impressive I thought.
http://i528.photobucket.com/albums/dd325/senecat/again/Screenshot2010-02-02at14100AM.png
I would think there's enough data there for them to work with, if they decide to.
So I just wanted to tell you and the others here this little story as part of the ongoing saga, knowing full well that where it will lead is probably nowhere. But I'll post back if they do, so we'll see.
Hope your doing well, and that you've gotten some rest. And yes, I believe you are the ideal person to contact Ars Techica if that was legit. Getting them to report on this would be a real coup.
Also, I keep seeing someone say the maximum temp is 70C for our machines. It's not, both Apple and Hardware Monitor say it's 95C. I don't want to argue that though, I just want people whose temps hit like 65-70C to know they can relax a bit if temps are their main concern. It's still a concern and so is the rest that has been written about all of this, but for me and a few others whose core temps can sometimes hit 90C , our concern there is pretty critical.
Take care, bud
Steve, Sam, whoever...
No4mk2 (Steve),
Nice work continuing the push against Apple. I'm hesitant to even tell you this as I don't won't to spoil your seemingly positive experience with Applecare today, but the Capture Data step is a dead end. They sent me down that path over a month ago as well, and after a lengthy period of silence, I was advised that somebody at Apple product engineering analyzed my data, and was happy to report that it is all "within limits". You can imagine how overjoyed I was to hear that. The fact that they are even making you run this app on your machine when they could just as easily reproduce the entire fault on one of their own machines pretty much sums up their total failure to listen to you / us.
Regarding the temperature limit (70C vs 95C), I think we need to clarify which temperature we are talking about. I have a feeling you are talking about Core temps. You are right when you say these core temps can approach upwards of 95C, but with that said, what Intel has actually specified in their own Thermal Specifications Document is a maximum Tcase value as opposed to Tcore. This can all get pretty confusing, but the simple explanation is as follows. Intel specifies a maximum Tcase temperature (67.9C for Xeon W3540). This temperature can not be accurately measured with on-chip sensors, but it is a fair assumption that it falls somewhere in between the CPU Temperature Diode value and the CPU Heatsink value. Peoples actual temps may vary depending on ambient air temp, airflow (system under desk vs in open air), etc., however, it is clear from my own observations and that of others, that our Tcase temps are likely approaching this level by simply playing audio. Where this gets scary is when you throw some work into the equation. Doing any sort of processor intensive work with audio playing causes CPU Temperature Diode temps in the high 80C range on my system, and I estimate the Tcase temperature during these tests is at least 10C higher than Apple's limit. This can be somewhat confirmed by the kernel panics I have observed when running stress test apps.
Sorry for rambling.. Intel makes it pretty clear that temperatures at or above max Tcase will reduce the life of the processor, and may cause total failure. Not only has Apple inadequately engineered the fan logic of these machines, but they also shipped it with poorly written audio and FW drivers which cause this temperature / power spike. It seems that two separate issues are actually conspiring against our machines, and I don't like it at all.
Again, sorry if I have rained on your positive experience with Applecare, but I don't have a lot of confidence that your case will end any differently than mine did (5 times).
Nice work for trying though! I will try and contact Ars Technica, and on a positive note, I did get a response from The Consumerist today. They are very interested!
Regards,
The deaf but not dumb claim ~100C
Intel seem to specify ~70C
but you have to be clear about where you are reading your temperature.
I go by Temperature Monitor's CPU A Temperature Diode, the heat sink will measure a lot less (& iStat Menus reading) and the CPU Core readings ~10C higher. smacman worked out and translated all this earlier in this thread.
See Intel specs re. temperature tolerances
http://www.intel.com/Assets/en_US/PDF/designguide/321323.pdf
From Table E.1 (p. 87) max sustained temp:
51.9 + 60*0.302 = 70C
(but I hope to be corrected and clarified by the competent)
Did somebody get out of the wrong side of the bed?
This could indicate that Apple have in fact fixed the issue in new Mac Pros, if they can replicate their results (ie: no loss of performance) in some more tests. Why would you be so against this possibility? Frankly, that would be be valuable data to know.
Sure, Apple need to fix the machines that have the problem, but if somebody has just bought a new Mac Pro that doesn't appear to have the issue, why are you so hostile? Jealous that your machine is faulty and you've had to waste so much time?
It is more that we've have people pile in with righteous indignation, taking a stance designed to defend holy Apple, rather than look at what their machine is actually doing. They tend to be loud and combative, but when the penny drops and they see what is going on they slink away quietly and without setting the record straight.
The fear is that it is exactly this which Apple support people do. Look for seconds only, with preconceived ideas and a minds already made up.
I am absolutely sure that if an modified MP system came out with a fix for this bug, smacman, me and all would be encouraged, and look for ways to use this knowledge to angle for a fix for all. It, at the least, would infer that Apple is capable of fixing it, if not why they don't issue a fix for us.
Methinks that all the heavyweights here who've been wresting this bug for so long were to hear of a genuine case of it not occurring, they'd be buttering up the poster in hope of getting cooperation to do comparative tests.
But some of us feel annoyed by what seem like the ambition of ego-bigot types of post, not that they are going to get anywhere, as the truth will out eventually. Eventually we'll publish a wall of shame for all those whose Apple worship got so out of hand and to make them work in destructive ways. It's the web era now, everything digital can and will be used in evidence against you.
This inappropriate vigilante mood is brought on by the Apple silence game, but at least they haven't tried the "no bread, let them eat cake" windup yet (in this case, burning bug - take a cold bath).
Guys, I had a vision - I see the PC guy and the Mac guy in the white room. In front of them is my 09 Mac Pro. The PC guy says "Hey, have you seen this fabulous piece of hardware? It runs Windows 7 even better than Snow Leopard..."
I saw Steve Jobs recorded graduation speech at Stanford. At his roots, he's a humble guy who started out with nothing and built an empire. Apple's reaction to this issue shows none of those qualities - only arrogance and a reaction fitting for a company more concerned about admitting liability than helping folks who support them. The fact that few web sites, if any, have shown any desire to stick their neck out and perform some real journalism speaks to the reach and influence of Apple. I now have contacted three main tech sites - none has shown any desire to investigate - I believe out of fear of reprisal from Apple.
Apologies for the early morning rant - the thought of investing over $5000 to switch from a PC to a Mac has me upset this morning....
I saw Steve Jobs recorded graduation speech at Stanford. At his roots, he's a humble guy who started out with nothing and built an empire. Apple's reaction to this issue shows none of those qualities - only arrogance and a reaction fitting for a company more concerned about admitting liability than helping folks who support them. The fact that few web sites, if any, have shown any desire to stick their neck out and perform some real journalism speaks to the reach and influence of Apple. I now have contacted three main tech sites - none has shown any desire to investigate - I believe out of fear of reprisal from Apple.
Apologies for the early morning rant - the thought of investing over $5000 to switch from a PC to a Mac has me upset this morning....
I thought about something about my patch:
I mentioned it disables Turbo and that makes you lose performance.
Well, Turbo is disabled anyway when CPU is running hot. So if you are getting high temps using your 2009 Mac Pro, turbo will turn of anyway.
Basically what this means is that the performance loss some people where mentioning when audio is decoding is probably because Turbo turns off as a result of high temps.
Furthermore this means that when using my patch on a system where you are experiencing high temps, will not decrease performance any more then what audio decoding is already doing.
So if you are having this temp problem all the time, the viability of my fix just raised. The only thing you will lose is ~8c higher idle temps.
I mentioned it disables Turbo and that makes you lose performance.
Well, Turbo is disabled anyway when CPU is running hot. So if you are getting high temps using your 2009 Mac Pro, turbo will turn of anyway.
Basically what this means is that the performance loss some people where mentioning when audio is decoding is probably because Turbo turns off as a result of high temps.
Furthermore this means that when using my patch on a system where you are experiencing high temps, will not decrease performance any more then what audio decoding is already doing.
So if you are having this temp problem all the time, the viability of my fix just raised. The only thing you will lose is ~8c higher idle temps.
So a 5% decrease in clock speed = 25% less performance?
Sorry but no
I'll prove you wrong in a short moment!
*wacks fans to maximum*Ok here we go:
Cinebench OpenGL test:
Without audio: 6386
With Audio: 5186
18.9% loss of performance, max CPU core temp of 45C, max heatsink temp of... 31C, max CPU load of... 12%.
Turbo mode wouldn't kick in under those condtions as the machine is hardly loaded.
There we go.