Optimum Temps for Mac Pro

[usasalazar]

Hi -

Anyone running smcFan control care to share their thoughts and ideas for best fan speeds (RPM) and temps for Intel MacPro?

Thanks!

 

[AZREOSpecialist]

I have a 3.32 GHz Core i7 in my Quad, but temps were 38-40C before using SMC Fan Control and are now down to 28-30C using SMC Fan Control. I basically doubled the minimum fan RPMs and the CPU and RAM have cooled by 10C.

 

[bobpensik]

Same as the person above, I have found that basically doubling the default speeds on my 2008 Mac Pro has significantly decreased the temperature. HOWEVER, it has also made the machine A LOT louder, and I am not sure I really like the trade off, but then again, even at default fan speeds my Mac Pro is pretty dam loud. I miss how quiet my laptop was!

 

[jwt]

29C surfing, 32C playing HD video on the 2.8 SP Harpertown

 

[peskaa]

8x2.8 2008 Harpertown

 

[Tesselator]

Hi -

Anyone running smcFan control care to share their thoughts and ideas for best fan speeds (RPM) and temps for Intel MacPro?

Thanks!

Well, there's a lot to consider here but the whole thing cam be simplified to Component safety and desired temps. Here's a good rule of thumb:

CPU's: Always below 55C please!
RAM: Always below 60C please!
HDDs: Always below 45C please!

And those are the critical components to watch and basically the desired temps across all models and makes. There are some exceptions but they don't vary very much. If you notice temps getting higher than any of these then create an smcFanControl preset for that situation and select it when you're doing those things.


Here's a pretty good post that summarizes some of the thoughts and many of the mechanical details about Mac Pro's fans, smcFanControl, and common sense temperatures:

• https://forums.macrumors.com/posts/7491799/

 

[nanofrog]

Well, there's a lot to consider here but the whole thing cam be simplified to Component safety and desired temps. Here's a good rule of thumb:

CPU's: Always below 55C please!
RAM: Always below 60C please!
HDDs: Always below 45C please!

And those are the critical components to watch and basically the desired temps across all models and makes. There are some exceptions but they don't vary very much. If you notice temps getting higher than any of these then create an smcFanControl preset for that situation and select it when you're doing those things.


Here's a pretty good post that summarizes some of the thoughts and many of the mechanical details about Mac Pro's fans, smcFanControl, and common sense temperatures:

• https://forums.macrumors.com/posts/7491799/

Generally, I agree. I have noticed however, Nehalem runs a little hotter (load).

My stock temps idle at the same level AZREOSpecialist's model did at stock SMC settings, and I don't know what that system did before/after for load (I'm assuming the values posted are idle, as it doesn't run that cool loaded ). But I'm getting 67C at 100% on all 4 cores. Tad warm, but not critical, given Intel allows them to hit 100C (actual value varies, and Intel won't state it). I panic if it hits 80C, and would make immediate changes. The hottest it ever got was 77C, and it made me too nervous, so I dialed it back a little (load hovered at 72C).

I've not yet OC'd the system again, to see what it does with the higher voltages and BCLK settings.

 

[Tesselator]

Generally, I agree. I have noticed however, Nehalem runs a little hotter (load).

My stock temps idle at the same level AZREOSpecialist's model did at stock SMC settings, and I don't know what that system did before/after for load (I'm assuming the values posted are idle, as it doesn't run that cool loaded ). But I'm getting 67C at 100% on all 4 cores. Tad warm, but not critical, given Intel allows them to hit 100C (actual value varies, and Intel won't state it). I panic if it hits 80C, and would make immediate changes. The hottest it ever got was 77C, and it made me too nervous, so I dialed it back a little (load hovered at 72C).

I've not yet OC'd the system again, to see what it does with the higher voltages and BCLK settings.

OK, make that 70C for the new 5500 series processors then. Keep in mind though than even Intel gets very nervous at 75C and from that point the processor WILL go into a TCC partial shut-down. 60C ~ 65C is still a better target (of course) and should be obtainable if there's a good heat-sink and a user controllable fan. For details see:

• https://forums.macrumors.com/posts/8022325/

 

[AppleWorking]

I panic if it hits 80C, and would make immediate changes. The hottest it ever got was 77C...

Hi nano, I figure you mean a temp of 80c in Windows, right?

Anyway, here's my 2 cents on this... When running 3DMark Vantage in Windows my CPU hits a max temp of 76c for a second or two but it never goes over with my current fan setting. I would prefer it stay under 75c but it's not possible if I want to keep my MP as silent as I like it in Vista... I find in OSX it's a different story, I never push my MP over 65c (ever) since OSX seems to measure the heatsink temp and not the core temp. This is why I don't really believe what I see using istat in OSX. I would hate to be close to "TJ Max" and not know about it.

To control my fans in windows (bootcamp not virtual) I first boot into OSX, switch profiles with smcFanControl and then reboot into vista using bootchamp (broken in SL). The reason is because the 2009 MP fans do NOT ramp up in vista. It's pretty much a sure fire way to fry your computer, since this is Apple's undocumented "feature" of bootcamp. I don't know if the new bootcamp in snow kitty solves this, but I don't really think it's fixed knowing Apple.

Anyhoo, these are my fan speeds in Vista:

Cooling Fans
Fan #1 1239 RPM
Fan #2 887 RPM
Fan #3 1217 RPM
Fan #4 1217 RPM
Fan #5 1817 RPM

OSX equivalent: #5 is BOOSTA, #3 & #4 Intake/Exhaust, #2 PS, #1 PCI

To control my fans in OSX I use this below. It's a awesome little application that Concorde Rules modded from an old program. Works like a charm in OSX on a 2009 MP. IMO, it's a billion times better than smcFanControl because fans are adjusted dynamically. You don't have to switch between profiles (like you have to with smcFanControl) the fans will go up and down as needed. Plus you won't be abusing your fans by leaving them ramped up as a constant in OSX.

https://forums.macrumors.com/threads/738163/

Please thank this guy for this, to me he's not getting the credit he deserves!

 

[Tesselator]

https://forums.macrumors.com/threads/738163/

Please thank this guy for this, to me he's not getting the credit he deserves!

He doesn't need credit. He needs someone to help him as indeed he's asked for directly more than once already. He's got the source code to someone else's program and he's floundering a bit. In it's current state it doesn't work.

 

[nanofrog]

OK, make that 70C for the new 5500 series processors then. Keep in mind though than even Intel gets very nervous at 75C and from that point the processor WILL go into a TCC partial shut-down. 60C ~ 65C is still a better target (of course) and should be obtainable if there's a good heat-sink and a user controllable fan.

I just went back and looked, and at spec, they list Tcase = 67 or 67.9C. So 70C is close enough, but a little cooler is better.

Examples:
i7-975
i7-920 (C0 stepping)
W3580
W5580 (DP example)

In the case of the units used in the '09 MP's, it's Tjunction, and not listed. But the above data should be an idea of what to expect in an ideal situation. Unfortunately, it's not that easy to do, as I'd guess those temps where determined on a bench, and in open conditions (not stuffed in a computer case). So the variance would make sense, assuming my recollection of 100 - 108C is accurate, as real world conditions rarely coincide with the benchies.

As per the 100 - 108C limit, I just can't find the source.

Hi nano, I figure you mean a temp of 80c in Windows, right?

Yes.

To be specific, it was Win7 RC 7100-64 bit.

Anyway, here's my 2 cents on this... When running 3DMark Vantage in Windows my CPU hits a max temp of 76c for a second or two but it never goes over with my current fan setting. I would prefer it stay under 75c but it's not possible if I want to keep my MP as silent as I like it in Vista... I find in OSX it's a different story, I never push my MP over 65c (ever) since OSX seems to measure the heatsink temp and not the core temp. This is why I don't really believe what I see using istat in OSX. I would hate to be close to "TJ Max" and not know about it.

I understand.

Even temps taken off the CPU diodes aren't accurate (10 - 15C variance on newer processors), and need to be measured with a thermocouple. You then use the measurement to make a calibration adjustment (offset) to get accurate temps off the reading. The setting, if possible, is done in the temp software. Real Temp 3.0 (windows) is one such example that allows you to do this.

The thermocouple is mounted dead center, and it would give you Tjunction, as it's sans IHS in the '09 MP's. As this isn't to be done without a heatsink (IHS or not), you need the thermocouple centered in a measured and machined copper plate, and TIM on both sides. Not an easy thing to come by. It usually has to be made for the specific package (processor).

So temp readings aren't accurate anyway, but a change is. That's due to the differences between real and displayed are negated when the difference is determined on the same system. You may not have accurate min & max, but delta T is still be useful, as it's at least accurate.

 

[AppleWorking]

He doesn't need credit. He needs someone to help him as indeed he's asked for directly more than once already. He's got the source code to someone else's program and he's floundering a bit. In it's current state it doesn't work.

In it's current state it DOES work perfectly. Have you tried it?

He merely wants to improve its current working state, that's all, and with this he is having no luck. But other than this it works very well, and far better than switching profiles constantly and keeping track of what your current temps are as if one has nothing better to do.

Just my opinion, because I do understand that enthusiasts love to switch profiles and have TOTAL control over their fans. I too watch my fan speed and temps even though his application works without user interaction, except for setup.

I believe in its current state - the boosta fan is linked to intake and exhaust. PCI works independently, just as does PS - and what he would like to do is separate the control of each into zones, and/or group them into zones. I don't feel this is necessary, but he does. To me, it works perfectly.

 

[AppleWorking]

I just went back and looked, and at spec, they list Tcase = 67 or 67.9C. So 70C is close enough, but a little cooler is better.

Examples:
i7-975
i7-920 (C0 stepping)
W3580
W5580 (DP example)

In the case of the units used in the '09 MP's, it's Tjunction, and not listed. But the above data should be an idea of what to expect in an ideal situation. Unfortunately, it's not that easy to do, as I'd guess those temps where determined on a bench, and in open conditions (not stuffed in a computer case). So the variance would make sense, assuming my recollection of 100 - 108C is accurate, as real world conditions rarely coincide with the benchies.

As per the 100 - 108C limit, I just can't find the source.

I totally agree with this even though I don't know as much as you. I always believed the max temp before the CPU throttles itself was around 98c no matter what Intel reported. Although I had no idea under what conditions they were making their recommendation... But now I kind of know (thank you), since you say they make this determination on a bench. I see. You have to admit though there are some individuals who look at Intel data and think it is written in stone without real world testing... (i wonder who that is, lol)

 

[nanofrog]

I totally agree with this even though I don't know as much as you. I always believed the max temp before the CPU throttles itself was around 98c no matter what Intel reported. Although I had no idea under what conditions they were making their recommendation... But now I kind of know (thank you), since you say they make this determination on a bench. I see. You have to admit though there are some individuals who look at Intel data and think it is written in stone without real world testing... (i wonder who that is, lol)

I keep looking for the throttle temp information, but I keep hitting a wall with it.

Board makers can set their own temps as well. At least this is done on better boards. I just checked the board I'm using, and I can't find the temps they set either. //grr

 

[AppleWorking]

Board makers can set their own temps as well. At least this is done on better boards. I just checked the board I'm using, and I can't find the temps they set either. //grr

Hi nano, I didn't know that. So board makers can actually override when a CPU will throttle itself back? That's interesting, I assume that would be a real plus for someone overclocking their chip. I wish we could overclock the new CPUs in the MP. And imagine further if Apple prevented the CPU from throttling down through their MB, where we could push our CPUs to the envelope of thermal meltdown. Wow, we would be in heaven. lol

I know that's about as likely as stretching a gnats behind over a ten gallon bucket, but it sure would be nice if apple sold a overclocking rig for fun as a limited edition. I think they would sell out in .2 seconds.

 

[nanofrog]

Hi nano, I didn't know that. So board makers can actually override when a CPU will throttle itself back? That's interesting, I assume that would be a real plus for someone overclocking their chip. I wish we could overclock the new CPUs in the MP. And imagine further if Apple prevented the CPU from throttling down through their MB, where we could push our CPUs to the envelope of thermal meltdown. Wow, we would be in heaven. lol

I know that's about as likely as stretching a gnats behind over a ten gallon bucket, but it sure would be nice if apple sold a overclocking rig for fun as a limited edition. I think they would sell out in .2 seconds.

ZDNet in germany created a utility that allowed OC'ing on the earlier MP's.

Perhaps someone will make a 3rd party utility that can do it for the '09's. Other than that, it would mean a modification of the chip itself, similar to BSEL mods on the Penryn based parts. Assuming it's even possible this time around.

 

[AppleWorking]

ZDNet in germany created a utility that allowed OC'ing on the earlier MP's.

Yup, heard about that.

Perhaps someone will make a 3rd party utility that can do it for the '09's.

Crosses fingers.

 

[nanofrog]

Yup, heard about that.

Crosses fingers.

I also went and looked to see if anyone's found a physical moddification for non OC compatible boards. Unfortunately, nothing came up.

With the boards so readily available for the i7 performance/Xeon 35xx (workstation parts), I don't think it will be so easy to find out if it's been done. Less motivation to try, and left to a much smaller group to figure it out. Namely the Xeon 55xx DP board users. At best, I think it will take awhile, if it even happens.

 

[Tesselator]

In it's current state it DOES work perfectly. Have you tried it?

Not the newest version no. I'm going by what he said. He said it's buggy and not ready. I guess he would know.

Also short of changing out the motherboard for something non-apple it will NOT be possible to control the fans individually nor take control over the PSU fan. All of the fans except the PSU fan have a hardware dependancy on at least one other fan in the system. That's on a hardware level. There's no getting around it AFAIK.



.

 

[nanofrog]

Not the newest version no. I'm going by what he said. He said it's buggy and not ready. I guess he would know.

Also short of changing out the motherboard for something non-apple it will NOT be possible to control the fans individually nor take control over the PSU fan. All of the fans except the PSU fan have a hardware dependancy on at least one other fan in the system. That's on a hardware level. There's no getting around it AFAIK..

You don't think a hardware fan controller would work?

 

[Tesselator]

You don't think a hardware fan controller would work?

A hardware fan controller? You mean unplug the fans from the MB and plug them into some card or separate controller? Is there such a thing?

The fans on the Apple MB aren't completely independent. You can't increase the voltage to one without affecting at least one other.

Here's the actual relationships:

 

[nanofrog]

A hardware fan controller? You mean unplug the fans from the MB and plug them into some card or separate controller? Is there such a thing?

Yep.

First link to pop up in a search. Many more were that came from.

The fans on the Apple MB aren't completely independent. You can't increase the voltage to one without affecting at least one other.

Here's the actual relationships:

This is actually quite common. Most boards only have 2 - 4 channels for fans that are independent of one another (not for the CPU/s), and multiple fans can be controlled by a single channel. Typically up to 3. Check out various PC boards (maybe have to look at the manuals), but it's in there. Despite the EFI firmware, it's still a hardware limit, and the board is made by Intel (design + manufactured).

BTW, your fan profile won't be the same as the '09's, but should be for '06 - 08's, as little changed.

 

[Tesselator]

Yep.

First link to pop up in a search. Many more were that came from.

Sure those would work - if we had a place to put them that is. :S

This is actually quite common. Most boards only have 2 - 4 channels for fans that are independent of one another (not for the CPU/s), and multiple fans can be controlled by a single channel. Typically up to 3. Check out various PC boards (maybe have to look at the manuals), but it's in there. Despite the EFI firmware, it's still a hardware limit, and the board is made by Intel (design + manufactured).

Yep.

BTW, your fan profile won't be the same as the '09's, but should be for '06 - 08's, as little changed.

Yeah, I figured so.

 

[nanofrog]

Sure those would work - if we had a place to put them that is. :S

There's much simpler ones, like this. It's just a ladder network of resistors tied to a switch (3 voltage dividers).

Surely these can fit. They're about the size of a large postage stamp.

 

[AppleWorking]

Not the newest version no. I'm going by what he said. He said it's buggy and not ready. I guess he would know.

Also short of changing out the motherboard for something non-apple it will NOT be possible to control the fans individually nor take control over the PSU fan. All of the fans except the PSU fan have a hardware dependancy on at least one other fan in the system. That's on a hardware level. There's no getting around it AFAIK.

Nah, it's ready. Especially since he can't remove the fan dependancies, it is even more ready than he thinks. LOL