Perhaps the best way to approach this with Apple is to complain that simply connecting a peripheral (eg. an audio interface) causes my Mac to consume more power, temps to rise, and performance to drop.
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
Is that limited to just audio periphery, or in general?
I use eSATA and Firewire drives and as far as I remember, that doesn't result in considerably higher power consumption and higher temps. I'm not sure though and will test that when I'm back at my Pro.
I use eSATA and Firewire drives and as far as I remember, that doesn't result in considerably higher power consumption and higher temps. I'm not sure though and will test that when I'm back at my Pro.
Constant writing (large files or folders) to a Firewire drive will also trigger this. The spike however only occurs while the disk is being written to, so unless you are doing a large transfer, it won't be evident.
Add "If using audio causes temperatures to rise much above 70C, shut the machine down"
Err duh, how am I meant to be monitoring the temperatures 24/7 ...
and why should I have to stop working when doing such an innocuous things as evoking core audio (NB. I have no control over what applications might use it and when).
It is clear that a combination of their attitude and training has made Apple tech support, on this issue, worse than useless and a complete waste of time.
If any support people have ever tried to report the issue, it seems obvious that Apple central are impervious to it. At the moment there seems nothing but brick wall.
I will never consider buying another Mac Pro unless Apple fix this issue on the 2009 systems, and soon. Not because the machine is a lemon re. the audio bug, but because my confidence in Apple's capability is completely undermined by their intransigence. I begin to suspect that this bug it is beyond Apple's capability and their policy is some sort of cover-up.
By my feet are 5 generations of pro Macs, and it has come to this.
You have to realise that most probably this is a software bug, and not Intel hardware issue. Hardware works fine under Windows.
Bashing MacPro development team won't help.
In the unlikely case of a CPU problem, it can be possibly solved by latest microcode update which can be [in theory] loaded from a kernel extension.
Bashing MacPro development team won't help.
In the unlikely case of a CPU problem, it can be possibly solved by latest microcode update which can be [in theory] loaded from a kernel extension.
Just received the official answer from Apple: "Your system seems to be absolutely within operational limits. We are working on further investigation on the issue" (free translation of the italian answer received).
I have to add that using an external FW hdd don't trigger the problem in our MP. I tried to copy several huge files (4-8 Gb) and cpu's temperature shows no significative variation
I have to add that using an external FW hdd don't trigger the problem in our MP. I tried to copy several huge files (4-8 Gb) and cpu's temperature shows no significative variation
Just wanted to report that this morning I watched a full length .avi movie via Divix plug in. After an hour my CPU was at 47C/117F - temps much lower than those observed during iTunes playback.
In addition, the source file for the movie was on a USB thumb drive, if that holds any significance.
CPU temp provided via Temperature monitor...
In addition, the source file for the movie was on a USB thumb drive, if that holds any significance.
CPU temp provided via Temperature monitor...
This is how I see it. If it was a hardware glitch, it would show up under Windows as well, which isn't happening. That leaves OS X as the source of the issue.
The only hardware aspect that might be able to stand some adjustment, is the SMC code, which can be done in an update.
smacman,
Im working on a little CPU fan speed control app that adjusts depending on temps.
Interested?
I started the project a while ago but since going to Snow Leopard I've dropped it but 95C load isn't very nice
Cheers
Im working on a little CPU fan speed control app that adjusts depending on temps.
Interested?
I started the project a while ago but since going to Snow Leopard I've dropped it but 95C load isn't very nice
Cheers
smacman,
Im working on a little CPU fan speed control app that adjusts depending on temps.
Interested?
I started the project a while ago but since going to Snow Leopard I've dropped it but 95C load isn't very nice
Cheers
Something like FanControl ?
It would make the designed to be quiet MacPro sound like systems based on the high-heat Pentium IV CPU's.
Sorry if this has already been addressed...
I just ran a quick test of playing various songs in iTunes for 20 minutes with the visualizer and equalizer enabled, then simultaneously played a streaming 1080p movie trailer available on www.quicktime.com. My CPU A Temperature Diode went from 32C to 51C while playing iTunes, then to 53C while playing the streaming HD movie. Back to 51C after closing the movie trailer, and quickly down to 34C shortly after quitting iTunes. When CPU A registered 53C, the CPU A Heatsink registered 43C, or 10C less than the CPU value. Core temps are about 10C higher than the CPU temp, bouncing between 62C and 63C while the CPU A Temperature registered 53C.
The maximum case temperature for my W3580 is 67.9C according to Intel's data sheet (Tcase_max at 130W). Judging by what I see in Temperature Monitor, my case temps don't go above 53C while listening to iTunes w/ visualizer on and playing back a 1080p movie trailer. As listed in this interesting thread, the core temps can get quite a bit higher than the case temp. Perhaps people are confusing the maximum case temp value of 67.9C with the readings they are getting off the actual cores?
While my CPU case temps rise approx. 20C above idle w/ iTunes and playing a movie trailer, at least on my machine it doesn't put me anywhere close to the maximum 67.9C. The Xeons and Core series have a built-in switch to shut down the cores and/or CPU in the event temperatures exceed thermal limits. So your CPU will know much earlier than you if it's in dangerous territory and start shutting itself down as needed to prevent damage.
I just ran a quick test of playing various songs in iTunes for 20 minutes with the visualizer and equalizer enabled, then simultaneously played a streaming 1080p movie trailer available on www.quicktime.com. My CPU A Temperature Diode went from 32C to 51C while playing iTunes, then to 53C while playing the streaming HD movie. Back to 51C after closing the movie trailer, and quickly down to 34C shortly after quitting iTunes. When CPU A registered 53C, the CPU A Heatsink registered 43C, or 10C less than the CPU value. Core temps are about 10C higher than the CPU temp, bouncing between 62C and 63C while the CPU A Temperature registered 53C.
The maximum case temperature for my W3580 is 67.9C according to Intel's data sheet (Tcase_max at 130W). Judging by what I see in Temperature Monitor, my case temps don't go above 53C while listening to iTunes w/ visualizer on and playing back a 1080p movie trailer. As listed in this interesting thread, the core temps can get quite a bit higher than the case temp. Perhaps people are confusing the maximum case temp value of 67.9C with the readings they are getting off the actual cores?
While my CPU case temps rise approx. 20C above idle w/ iTunes and playing a movie trailer, at least on my machine it doesn't put me anywhere close to the maximum 67.9C. The Xeons and Core series have a built-in switch to shut down the cores and/or CPU in the event temperatures exceed thermal limits. So your CPU will know much earlier than you if it's in dangerous territory and start shutting itself down as needed to prevent damage.
For some reason, you are seeing much cooler temps than me. If I do similar tests, my Heatsink temp reaches 65C, CPU Diode reaches 75C, and the core temps can be well into the 80C range..
Question for AZREOSpecialist:
Your results got me thinking, and I seem to recall you did some CPU transplants in your machine. Is your current W3580 chip installed by you or did you order a new machine from Apple? I'm thinking that perhaps the reason your temps are so much lower, is that with your self installed CPU, you probably did a much better job applying thermal paste than Apple did.
In any case, I'm curious to see if your temps remain ok when your cores are working hard. Any observations? In my case, audio + core stress = kernel panic. It seems like when Intel thermal protection logic kicks in, OSX kernel panics. In Windows, I observed what I think was CPU throttling. When running a stress test, one core reached 99C, then suddenly all four cores cooled down to 85C and slowly worked back up to 94C. Either way, audio should not be causing any significant rise in temps in OSX.
Lastly, I am still a little grey on which sensor equates best to Intel's Tcase. Is it CPU A Heat Sink or CPU A Temperature Diode?
Your results got me thinking, and I seem to recall you did some CPU transplants in your machine. Is your current W3580 chip installed by you or did you order a new machine from Apple? I'm thinking that perhaps the reason your temps are so much lower, is that with your self installed CPU, you probably did a much better job applying thermal paste than Apple did.
In any case, I'm curious to see if your temps remain ok when your cores are working hard. Any observations? In my case, audio + core stress = kernel panic. It seems like when Intel thermal protection logic kicks in, OSX kernel panics. In Windows, I observed what I think was CPU throttling. When running a stress test, one core reached 99C, then suddenly all four cores cooled down to 85C and slowly worked back up to 94C. Either way, audio should not be causing any significant rise in temps in OSX.
Lastly, I am still a little grey on which sensor equates best to Intel's Tcase. Is it CPU A Heat Sink or CPU A Temperature Diode?
smacman,
Im working on a little CPU fan speed control app that adjusts depending on temps.
Interested?
I started the project a while ago but since going to Snow Leopard I've dropped it but 95C load isn't very nice
Cheers
I'm very interested! Ideally, we'd be able to keep the CPU A Temperature Diode below 67.9 under any conditions.
May I ask, have you tried to quantify the performance impact of
connecting the cable you mentioned before? E.g., something like
running Geekbench (which is CPU-biased) when the cable is attached
and when it is not and checking the results?
If there is hidden CPU usage, not revealed by Activity Monitor, say,
it might be revealed indirectly through its impact on other processes
that need to use the CPU.
tbh the data that has been collected here is quite conclusive in saying that there is no added CPU usage. it seems to all be QPI based.
if it was CPU based, we would surely be seeing this problem on the i7 quad core imacs.
I wouldn't like to speculate. No harm in checking those benchmarks, though.
We can be sure the CPU is doing something. It doesn't heat up just sitting
there
I wouldn't like to speculate. No harm in checking those benchmarks, though.
We can be sure the CPU is doing something. It doesn't heat up just sitting
there
that is true, i wasnt ruling it out - just stating that the data doesnt seem to indicate that the CPU is being utilised. very hard problem this is!
is there some sort of lower level application that can measure this sort of stuff?
Neither of them shows actual Tcase. Diode is on the die so it is separated by the silicon-IHS junction and heatsink sensor is separated by IHS-heatsink junction. So both temperatures are affected by the value of Rj*P where Rj is thermal resistance of the junction and P dissipated power so the temperature differential is not a constant value. In case of heatsink Rj is also dependent on quantity and quality of thermal paste application.
Using Tcores in control loop is the best approach however they are only available programmatically. If you were to restart a hot CPU it could crash long time before fan control drivers were loaded.
SMC would kick in immediately even during reboot because it works independently and therefore needs direct access to some sensor (heatsink.) Maybe using on-die diode is better still but isn't there a large sample-to-sample deviation in calibration values?
I know.. Hence my question.. Which one best represents Tcase? Since Intel specifies the CPU's absolute Max Temp as a Tcase value (67.9C), which reading will best approximate the true Tcase value? If it's the CPU Temperature Diode, then I think we are screwed. Almost every Quad is showing Diode temps in excess of 67.9 simply playing audio back. Stress the cores and it goes substantially higher even (80-90C on my machine).
The reason I am somewhat preoccupied by this at the moment, is so that the next time Apple tells me that the temp is "within limits", I can re-educate them.
Either of them can be used to estimate Tcase.
Tcase=Tdiode - Rdc * P
Tcase=Theatsink + Rch * P
Rdc = thermal resistance diode-case (K/W)
Rch = thermal resistance case-heatsink (K/W)
P = CPU dissipated power (W)
I have left out thermal inertia component so this assumes some stable conditions.
Interesting.. So for simplicity, Tcase is always somewhere in between the two values? What would typical Rdc and Rch values be?