Temperatures Question

[Soraka]

Hi everyone,

I'm new here, new to the bitcoining and everything.
Anyways, i had a question.

I have an avarage-ish computer, and i'm wondering about the temperatures of my computer.
I've been running the computer already for about 2-3 days straight, mining. Not non-stop, i sometimes shut it down when i need to work on something. On avarage 35 mhash, not ideal but it works quite okay.
My video card is a Nvidia GeForce 250 GTS, which runs at 75-80 degrees celcius on avarage now,
and my CPU is a Intel quad core2 Q6600, with an avarage temperature of 70.

I wanted to know if keeping the computer on at this temperature is too high or not, and if there are any ways i can lower the temperature if it's needed without changing the hardware.

Thanks for your time, hope to hear something.
Soraka

[1714146856]

1714146856

[1714146856]

1714146856

[1714146856]

1714146856

[1714146856]

1714146856

[jake262144]

You need to hold your horses there, cowboy. You're galloping head-over-heels into town while a black-hatted, shotgun-toting bad guy is camping on the saloon's balcony.
You need to do some reading first, get a grip on what's what here in the bitcoin corral ^^


MANY OTHERWISE REPUTABLE SOURCES MISINTERPRETED INTEL'S SPECS ON THE Q6000 CPU FAMILY.
Your Q6600 might be ready to go belly up if you're actually keeping it at 70 CELSIUS.
Maximum operating temp for this CPU is 60.3 to 62.2 °C ⁽¹⁾

I MISINTERPRETED THAT SAME SPEC AS WELL.
YOU'RE OK WITH THAT CPU TEMP SORAKA, EXCEPT...
Do you use this CPU for bitcoin mining? CPU bitcoin mining has not been a viable option for a very long time now.

While your 250 GTS is in the clear (maximum temp for this card is 105 °C ⁽²⁾), there is a snowball chance in hell you'll make ANY money with this card.
Its bitcoin mining performance is just abyssal - a 150 Watt card averaging 35 MHash/s is laughing stock.

You're wasting far more resources on power than you can earn with your mining. Not to mention you seem to be killing your machine in the process.

I don't mean to put you off but in order to actually make any kind of profit mining bitcoin these days, you need to select your hardware carefully.
Damn, with current bitcoin proces, even GPU mining seems to be on its last leg...

Make sure to visit this site to compare various devices:  
   https://en.bitcoin.it/wiki/Mining_hardware_comparison

A quick glance will tell you that a single Radeon HD6950 card is EIGHT times as energy efficient as your 250 GTS.
A very inexpensive HD6770 with a power draw similar to your GPU should be capable of 200 MHash/s easily.

Put in some time, browse the forums, do the research, and for Chrissakes please do something about your CPU temp, will you?

Feel free to ask further questions, I'll respond when I'm able to.
Jake


References:
(1)
   http://HTTP://DOWNLOAD.INTEL.COM/DESIGN/PROCESSOR/DATASHTS/31559205.PDF
   http://www.cpu-world.com/CPUs/Core_2/Intel-Core%202%20Quad%20Q6600%20HH80562PH0568M%20%28BX80562Q6600%29.html
   http://www.hardwaresecrets.com/article/143/5

(2) http://www.nvidia.com/object/product_geforce_gts_250_us.html

EDIT: UPPERCASE

[P4man]

Your Q6600 is ready to go belly up if you're actually keeping it at 70 CELSIUS.
Maximum operating temp for this CPU is 60.3 to 62.2 °C ⁽¹⁾

I saw your links, but that has to be a misinterpretation. There is no way any CPU is going to suffer from 62C. More likely its that the heatsink has be to designed to keep the temperature delta below 62C, which will correspond to 80+C temperature depending on ambient.I have a Q6600 myself, with the stock cooler you can not keep it below 62C even at stock speed (Im getting substantially lower temps @3Ghz with a huge aftermarket cooler, but that is besides the point). The fact thermal throtteling doesnt kick in until 90+C is another clear indication that 60 or 70C is fine.

Agreed it makes no sense to mine bitcoins with it though. Try litecoin.


Edit: googled around a bit. See this:
On the temperature side, we noted via Intel TAT software after 15 minutes of 4x Prime95 and the indiscrete Intel CPU cooler delivered with our model of the QX6700, the following results: 69°C with the QX6700 and 66°C with the Q6600.
http://www.behardware.com/articles/651-2/intel-core-2-quad-q6600.html

Its unthinkable intel would ship their once flag ship cpu with a cooler thats utterly unable to keep the cpu below its design specs in an open rig case with low ambient temperatures.

[jake262144]

I saw your links, but that has to be a misinterpretation.

Is the manufacturer's specification wrong as well?
http://download.intel.com/design/processor/datashts/31559205.pdf

If he has the 105 W TDP version, 62.2 celsius it is.
If it's a later revision with 95 W TDP, the max temp is 71 °C
Being no psychic I can't know which version of q6600 he has. OP doesn't sound like a power user as well, so I used the safer of the two values.

I'm aware that a lot of folks perceive the q6600 as an overclock beast but I'm just supplying hard, technical data.
You overspec your CPU and all bets are off.

I changed my original post a bit, so as not to scare the living crap out of Soraka ^^

[P4man]

I saw your links, but that has to be a misinterpretation.

Is the manufacturer's specification wrong as well?
http://download.intel.com/design/processor/datashts/31559205.pdf

No, but your interpretation of it is wrong. First of all, Tcase is not what the internal thermal sensor measures, its the outside of the heatspreader (you have no sensors there).

Moreover, what you are looking at is specifications for the heatsink, its not the maximum safe cpu temperatures.  Look at those charts, do you think a 10W CPU will get damaged at 44C? Ambient could be that high! Those temperatures are delta's, nothing else makes sense when speccing a heatsink. The heatsink has to be able to keep the temperature difference at those temps.

Again, do you think intel would not design its own heatsinks to conform to their own specs?

[Soraka]

Hello again,
Sorry for this late reply. I'd like to thank you all.
For the version of my CPU, i have this one : http://ark.intel.com/products/29765/Intel-Core2-Quad-Processor-Q6600-%288M-Cache-2_40-GHz-1066-MHz-FSB%29

I'm not really a power user, like you mentioned. It's a gaming rig that i have had made a few years back, and upgraded aswell.
For Jake, you cant really scare me. I've had things way worse to my older computers happen*cough managed to fry half the components cough*, currently on an IT course on university, so i can keep my head calm in most of those things.

Also, my computer is not overclocked. And before i started with mining, i've done a lot of research. I'm planning to upgrade aswell in the near future as my current specifications are starting to get too old. I'm not trying to get too much of a profit on Bitmining, it's that i always have my computer on for specific reasons. Mostly for remote control, and i figured i could let it do something if it's on anyways.

For the coolers. I have two at the moment. One is on the side of the case(regular case cooler). The other one is if i'm right a Ninja cooler. I cant see the exact model from where i am at the moment, but i will tell as soon as i am able to.

And for that link of mining hardware, i've already seen that during my research for everything

EDIT:
i forgot to ask. About the litecoins, i looked them up a bit. I dont assume there is any way Litecoins go over to Bitcoins other then selling on the market?

[P4man]

EDIT:
i forgot to ask. About the litecoins, i looked them up a bit. I dont assume there is any way Litecoins go over to Bitcoins other then selling on the market?

If you mean trade litecoins for bitcoins, just go here:
https://btc-e.com/

Just deposit your litecoins and sell them for bitcoins. Not gonna make you rich either, but at least it makes at least an order of magnitude more sense than mining bitcoins with your cpu.

[jake262144]

I don't mean to be rude, P4 but please try reading Intel's spec, ok?
These temps are NOT deltas!

Quoting from the aforementioned spec sheet
{
  In the event of a catastrophic cooling failure, the processor will
  automatically shut down when the silicon has reached a
  temperature approximately 20 °C above the maximum TC ⁽¹⁾. Assertion
  of THERMTRIP# (Thermal Trip) indicates the processor junction
  temperature has reached a level beyond where permanent silicon
  damage may occur.
}

Unless Intel has no clue, 82.2 (or 91, depending on the CPU revision ) °C is considered CATASTROPHIC COOLING FAILURE which may permanently damage the CPU.

Let me draw an analogy: do you think running a GPU at 110 °C is wise just because it's critical temp is 130 °C?

Nobody does that and lives happily ever after. Mind that bad guy on the balcony - and his two friends behind the sheriff's office as well ^^

Don't you run your GPUs at a temperatire radically lower than its TMAX?
Posts have been made here by guys who felt lucky and tried running their GPUs close to TMAX - they either RMA'd their GPUs or buried them in wooden boxes. Search them out if you so please.

Trying to keep everyone (and their hardware) safe,
Jake

(1) Yes, TC. Intel is using TCASE, not TJUNCTION here!

[jake262144]

You already told us you have a Q6600, Soraka. Ok.
There have been two versions (revisions) of this CPU: B3 (SL9UM) and G0 (SLACR).

G0 is the 95 Watt version while B3 is the earlier 105 Watt version.


To get this info, you need to either use a program like cpu-z or read the processor's serial number manually.
I'm aware that some (very few) BIOSes might display this info as well.

Soraka, can you download cpu-z ⁽¹⁾, run it, and post the results? The info you need is the "Revision" field.



(1)  http://www.cpuid.com/softwares/cpu-z.html

[Soraka]

Ah, seems like i messed up there. I thought that was with the revision aswell in the link that i gave. Guess i didn't read it well. Now i feel like an idiot, haha.

Now, lets see. The revision i have on the cpu is G0

Also, P4Man, thanks for that link. I'll keep mining for now, because i'm close to 0.1, and cash that in. Then i'll go for the litecoins. Thanks again

[P4man]

I don't mean to be rude, P4 but please try reading Intel's spec, ok?
These temps are NOT deltas!

Quoting from the aforementioned spec sheet
{
  In the event of a catastrophic cooling failure, the processor will
  automatically shut down when the silicon has reached a
  temperature approximately 20 °C above the maximum TC ⁽¹⁾. Assertion
  of THERMTRIP# (Thermal Trip) indicates the processor junction
  temperature has reached a level beyond where permanent silicon
  damage may occur.
}

Unless Intel has no clue, 82.2 (or 91, depending on the CPU revision ) °C is considered CATASTROPHIC COOLING FAILURE which may permanently damage the CPU.

So why doesnt it shut down at 82C? Because it doesnt, I have tested it. At 95C it starts throttling. I suspect it shuts down at 100C or more (which you can only achieve if you can disable throttling or put the damn thing in an oven). So indeed to get anywhere near those temps would have to involve a CATASTROPHIC COOLING FAILURE  as in a heatsink that fell off, there is no way to get that high otherwise.

You may be right those temps are not delta's, but the temps are tcase which are vastly different from the internal sensor and more importantly, they are specifications for the heatsink.

Let me draw an analogy: do you think running a GPU at 110 °C is wise just because it's critical temp is 130 °C?

No. Im actually quite anal about temps, my 5870 runs at 35C my other cards below 60C. But that doesnt mean I believe for one second that running any cpu at 60C is dangerous.

[P4man]

You will not begin to throttle because of temps until you are approximately 20C above Thermal Specification. For example, the Thermal Specification (Tcase) of the Q6600 G0 is 71C. At approximately 91C, it will begin to thermally throttle. This then makes sense why the TJ Max target is 90C; Intel is trying to make the sensor the most sensitive when it is at the point of damaging the CPU. TJ Max Target is not the point of damage, but it is close to the THERMTRIP# point (the point when the CPU automatically throttles and shuts down until proper temperatures are restored). Since as discussed earlier thermal specification (Tcase) is actually the temperature of the center of the IHS and it's temperatures are ussually 10C below core temps, there is actually a 10C buffer between Distance to TJ Max=0 and Permanent Silicone Damage. What Intel has done is said that they are not allowing for any difference between the temperature of the cores and the temperature of the IHS and thus we are aligning Tcase (without buffer) with TJ Max. Remember however that depending upon the solder the difference between Tcase and Core temps is ussually 10C, Intel is simply being safe and not allowing for any difference even if it exists.

If you look at most CPUs Tcase values, their TJ Max Targets are almost always 20C above their Thermal Specifications (Tcase) which perfectly matches with this concept.

http://www.overclock.net/t/476469/the-truth-about-temperatures-and-voltages

My Q6600 has a target TJ of 90C. Its not gonna melt at 70C.


http://intel.wingateweb.com/taiwan08/published/sessions/TPWS002/FA08%20IDF-Taipei_TPWS002_Nov_1006.pdf

[Gabi]

My Q9550 happily crunch on BOINC since i bought it (years ago) and it stay at like 60-65°C without problems.

70 is fine too anyway

The videocard temperature is fine too

The problem is that cpu mining is almost useless, you spend much more in energy than what you gain. And same apply for nvidia mining...

[axelj]

You will not begin to throttle because of temps until you are approximately 20C above Thermal Specification. For example, the Thermal Specification (Tcase) of the Q6600 G0 is 71C. At approximately 91C, it will begin to thermally throttle. This then makes sense why the TJ Max target is 90C; Intel is trying to make the sensor the most sensitive when it is at the point of damaging the CPU. TJ Max Target is not the point of damage, but it is close to the THERMTRIP# point (the point when the CPU automatically throttles and shuts down until proper temperatures are restored). Since as discussed earlier thermal specification (Tcase) is actually the temperature of the center of the IHS and it's temperatures are ussually 10C below core temps, there is actually a 10C buffer between Distance to TJ Max=0 and Permanent Silicone Damage. What Intel has done is said that they are not allowing for any difference between the temperature of the cores and the temperature of the IHS and thus we are aligning Tcase (without buffer) with TJ Max. Remember however that depending upon the solder the difference between Tcase and Core temps is ussually 10C, Intel is simply being safe and not allowing for any difference even if it exists.

If you look at most CPUs Tcase values, their TJ Max Targets are almost always 20C above their Thermal Specifications (Tcase) which perfectly matches with this concept.

http://www.overclock.net/t/476469/the-truth-about-temperatures-and-voltages

My Q6600 has a target TJ of 90C. Its not gonna melt at 70C.

http://i182.photobucket.com/albums/x31/chickeninferno/TJmax75percent.jpg
http://intel.wingateweb.com/taiwan08/published/sessions/TPWS002/FA08%20IDF-Taipei_TPWS002_Nov_1006.pdf

its not gonna melt, but have a processesor constantly at 70c makes it lifetime shorten a great time.. just so you know so yeah.. 70c is not dangerus but it will shorten the cpu life.

[P4man]

its not gonna melt, but have a processesor constantly at 70c makes it lifetime shorten a great time.. just so you know so yeah.. 70c is not dangerus but it will shorten the cpu life.

Lower is better, no one is arguing that. I just strongly disagree that the maximum operating temperature would be 60 or 62C as the above poster claimed. Intel puts the number somewhere between 95C and 105C.

Now it is true that higher temps will inevitably reduce lifespan. But so will running a CPU 24/7. Of course there is no  telling if that lifespan of your cpu is 10 years or 1000 years.

BTW, I dont like to rely on anecdotal proof, because it is no proof, but I do want to point out Ive had an Athlon 64 in my HTPC running near and often at its shutdown temperature for 2+ years. It was almost constantly above 90C and Ive recorded peaks of 135C (!) before it shut down and I had to reboot it.  

WHy would I do something insanely stupid like that? Because I was convinced the sensor was faulty;  this was a low power A64 with no reason to overheat and the heatsink never even got luke warm, fan speed didnt change a thing, nor did swapping coolers. I could even run the CPU with no heatsink and although it would shut down after just a few minutes, with the heat sensor registering 135C in the bios, the heatspreader was warm, but still not quite burning my fingers. So I decided the measurement sensor was faulty (and causing unnecessary shutdowns under load).

Little did I know I had a badly packaged chip and the CPU die did not make proper contact with the heatspreader, rendering the heatsink useless. The sensor was accurate, and the chip had no way to dissipate its heat. After more than 2 years, I removed the aluminum heatspreader, mounted a heatsink with a shim directly on the naked die and found the chip to be a stellar overclocker that has worked for years after.

Recommended? Most definitely not. But CPUs can stand a bit of heat.

[jake262144]

Having taken some time to dig through all of Intel's docs, I admit I may have been more concerned that necessary.

What's bugging me is, it seems that every single technology portal specifying the maximum temp for the q6600 made the same mistake I did:
opened up the latest version of Intel's specs -> thermal -> table 28 or table 29 depending on the TDP.
The ONLY thermal spec intel bothered to provide there is the TC value.
TC is the CPU temp alright, except it's measured on the IHS and hence lower than TJunction.

Since TC is usually about 10 °C lower than TJ, Soraka is at least 20 °C away from TAbsoluteMax.
He's in the green, he's a-ok.

My original post has been corrected.

Thanks for a great, technical discussion, and finally untangling this mess.

Hereby do I proclaim us Q6600 Experts.
P4 gets the Chief Q6600 Expert title.

BTW, your link seems to be dead P4 ( http://intel.wingateweb.com/taiwan08/published/sessions/TPWS002/FA08%20IDF-Taipei_TPWS002_Nov_1006.pdf )

[P4man]

Having taken some time to dig through all of Intel's docs, I admit I may have been more concerned that necessary.

What's bugging me is, it seems that every single technology portal specifying the maximum temp for the q6600 made the same mistake I did:
opened up the latest version of Intel's specs -> thermal -> table 28 or table 29 depending on the TDP.
The ONLY thermal spec intel bothered to provide there is the TC value.
TC is the CPU temp alright, except it's measured on the IHS and hence lower than TJunction.

You are right the same mistake is repeated everywhere, and I was confused by it initially myself; thing is, intel usually doesnt specify absolute maximum safe temperatures, because no one has to know. Throttling and termtip (with even margin on that) ensure you never get there anyway, as the chips are designed so they cant overheat. For "casual" overheating, say very high ambient temp + power virus,  throttling takes care of it. In the case of catastrophic failure, like mis-mounted heatsink, termtrip will still save your chip.

What they do publish in great detail is what the cooling solution has to be capable off to avoid throttling from ever happening in real world circumstances.

Since TC is usually about 10 °C lower than TJ, Soraka is at least 20 °C away from TAbsoluteMax.
He's in the green, he's a-ok.

My original post has been corrected.

Thanks for a great, technical discussion, and finally untangling this mess

Hereby do I proclaim us Q6600 Experts.
P4 gets the Chief Q6600 Expert title.

Im nothing of the sort; but common sense should tell you 62C cant be dangerous for a CPU.
I appreciate the response though.

BTW, your link seems to be dead P4 ( http://intel.wingateweb.com/taiwan08/published/sessions/TPWS002/FA08%20IDF-Taipei_TPWS002_Nov_1006.pdf )

try this one, has some good info on the mechanisms (as well as Tj target specs):
http://lenry.atw.hu/tjmax.pdf

[P4man]

BTW, slightly related, a long time ago I uploaded a video that shows throttling on a Pentium 4 prescott. Interesting is that it appears the thermal sensor on that system was rigged, and deliberately provided a much too low temperature when the CPU is not throtteling. It could be a bios bug, but it looks deliberate to me:
http://www.youtube.com/watch?v=_-Nx_gZ5VPU

Prescotts had a reputation for being notoriously hot, but anyone reviewing that system would think otherwise.