How room temp. affects processor temp. (graphs)

I was curious just how much room temperature can affect the temperature of the PC. I chose to look at the CPU since Speedfan can monitor and log temps for each of the 4 cores in my Q6600, and since of all the components, the CPU is probably the most sensitive to changes in air temp. since mine is air cooled.

Anyway, I did the same x264.exe encode that I have been doing for all my temperature comparisons and monitored room temp. with a calibrated digital thermometer at several points during each encode. These were averaged and graphed against the averaged reported core temp* values from Speedfan for the entire second pass of a 2-pass x264 encode of the same video file. I was happy to see that for the different room temps used, the increases were pretty linear (certainly within error).

Result:
You can see by the slopes of the regression line that every delta °F of room temp. affected the average core temp by about 0.8 °C and for your Celsius folks, every delta °C of room temp. affected the average core temp by about 1-1/2 °C.

So what does this mean and why do you care? Well, using these rules of thumb, if it's currently 70 °F in your room, and your average load core temp is 65 °C, you can expect that to change by roughly 0.8 °C for every single °F your room temp. change. Say your room hits 80 °F. Your load core temp should increase from 65 to 73 °C which may be unacceptable to you and you might want to adjust your o/c accordingly. This is just an approximation based on my system. Your mileage may vary...

*The numbers I used are equivalent to those collected by TAT or RMClock: these temps are core temps. As I understand it, TJunction never changes and is a fixed value for a given chip. The Quads get a values of 100 °C and the duals get 85 °C. The core temp is defined as:

Code:

---

Core temp = TJunction - DTS

Example, DTS reads 62.  You take 100-62=38 and your core temp is 38 °C.

---

DTS (Digital Thermal Sensor) can be read directly for each core. See this thread for more on reading your DTS directly without software like TAT or RMClock paying attention to uncleweb's posts using crystalCPUID to read the DTS directly. When I tried this method, I was able to get the same values for the core temps on my Q6600 as TAT and RMClock gave me. For some reason, Speedfan always shows cooler core temps for my chip which I corrected by adding 15 °C to each temp (the table is CORRECTED temps). Read more about that in the caption under the graph.

Raw data and graphs
Hardware specs: Q6600 (lapped) @ 9x333, Ultra-120 Extreme (lapped), P5B-Del., P182 case w/ 4 fans on low, Corsair 620HX, Ballistix DDR2-800 @ 4-4-4-12 (1:1 Mem:CPU).

http://img264.imageshack.us/img264/8686/tempsnewcu7.gif

Raw data table in case you want the individual points:

http://img264.imageshack.us/img264/8...tatablerb5.gif

Downloads and References
To download crystalCPUID: http://crystalmark.info/software/Cry...D/index-e.html
To download rmclock: http://cpu.rightmark.org/products/rmclock.shtml
To download speedfan: http://www.almico.com/speedfan.php
To download TAT: http://www.techpowerup.com/downloads/392/mirrors.php

To read more about TJunction and for a discussion about other things relating to thermal output fromC2D chips, see: http://forumz.tomshardware.com/hardw...topic&t=221745

Useful information, thanks for caclulating it out!
I notice from your specs you have the fans on low. Is there any chance you could repeat it with variable speeds to calculate how a delta Fan RPM affects the room:CPU temp ratio?
Thanks either way.

Wow great job!!! :)

Good info. The rise in component temps are definitely not proportional to the ambient temps in the room from what I've seen. Right now my E6600 @ 3.4ghz is idling at 30'c while earlier this week when we had that heatwave here in Southern California, it would idle around 42'c.

I don't have a way of reading temperature in my room which is the warmest of the house since my computer is like a space heater but I know that outside about 1/2 mile away at the Seal Beach Naval Weapons Station is only 16'c. (it's only 9:40am and slightly overcast)

This is good info for anyone interested. Room temperature makes a huge difference in CPU temps. My X2 runs 3-4c hotter on load in just a slightly warmer room. :flame:

Heh, I saw you post this at Anandtech.. and I replied with;

It's pretty much common sense. As ambient warms, so will everything else.. 1:1.

I guess this clears it up for some people though.

Hehe..

Quote:

---

Originally Posted by Renfield

I notice from your specs you have the fans on low. Is there any chance you could repeat it with variable speeds to calculate how a delta Fan RPM affects the room:CPU temp ratio?
Thanks either way.

---

That would be interesting since what you're really measuring in an experiment like that is how efficient your case airflow is... since you can't get cooler than ambient. It takes over 90 minutes per run with the file I'm using (I didn't use the temp data from the first pass though since it only uses about 80 % of the CPUs), and it was actually pretty tedious since I keep coming back to the room to record the temp on the digital thermometer by hand so I could average them. I don't think I'll be repeating this experiment with different case fan speeds, but as it heats up this summer, I'll probably record some higher room temp. results to see if that 1:2 ratio holds true.

Thanks for the interest.

Quote:

---

Originally Posted by Andrew LB

Good info. The rise in component temps are definitely not proportional to the ambient temps in the room from what I've seen. Right now my E6600 @ 3.4ghz is idling at 30'c while earlier this week when we had that heatwave here in Southern California, it would idle around 42'c.

---

I've seen the same thing. It was about 80 °F in here last week and my idle temps were like 10 °C higher than usual. What would be really cool would be sort of on-demand FSB sort of like what speedstep is to the multiplier for the FSB and the voltage. Maybe AISuite does this, I haven't messed with it.

The goal would be keeping the temps low, for example, if you're just going to be browsing the web or typing, you really don't need your o/c'ed values. I'd use this if I didn't have to reboot and reconfigure my BIOS each time I wanted to switch speeds and voltages.

Glad someone else is finding this useful -- always glad to contribute.

my room is extremely hot these days and my cpu idles 55c. It used to be in the mid 40s when it was colder.

Quote:

---

Originally Posted by kingdomwinds

my room is extremely hot these days and my cpu idles 55c. It used to be in the mid 40s when it was colder.

---

Dude, that's pretty hot for idle if those are tcase numbers...

Are they tcase or tjunction?
What are the load temps?
What are you cooling it with (HS) and how well ventilated is your case?

measured with TAT. My case is a centurion 5, which as 80mm intake and 120 exhaust. I use tuniq tower but i think i have bad contact.

So far, i have not had any stability problems but my load temps are 62c-65.

Today, it is not has hot as yesterday. So i am idling 47c right now.

Quote:

---

Originally Posted by kingdomwinds

measured with TAT. My case is a centurion 5, which as 80mm intake and 120 exhaust. I use tuniq tower but i think i have bad contact.

So far, i have not had any stability problems but my load temps are 62c-65.

Today, it is not has hot as yesterday. So i am idling 47c right now.

---

Those temps aren't too bad then but I'll bet you can easily shave off 6-10 °C of your temps by lapping both the CPU and HS. It's not as tough as you think it is was the conclusion I came to after doing it myself for the first time recently.

Here are the links to the threads I started here sharing my experiences with both of them; they have pictures and temp. results:

Q6600 IHS lapping
Ultra-120 Extreme lapping

Quote:

---

Originally Posted by NickS

Heh, I saw you post this at Anandtech.. and I replied with;

It's pretty much common sense. As ambient warms, so will everything else.. 1:1.

I guess this clears it up for some people though.

Hehe..

---

Someone who things shiney processors somehow help probably would post something like that.

Quote:

---

Originally Posted by falqon

Someone who things shiney processors somehow help probably would post something like that.

---

Im going ot give this one the benefit of the doubt and say he was sarcastic before I go and get banned for publically spanking him in front of everyone...

Ahh :banana::banana::banana::banana: it... :slapass: :slapass: :p:

yap well done ! now all buy air conditioning :D

yeah I've noticed it wasnt anywhere near a 1:1 increase a while ago and moved everything to the basement.

Window A/C units ftw.

So what are you saying here? If my room temperature goes up, so will the temp of my computer?!?!?! http://smiliesftw.com/x/hsugh.gif

Quote:

---

Originally Posted by Spawne32

So what are you saying here? If my room temperature goes up, so will the temp of my computer?!?!?! http://smiliesftw.com/x/hsugh.gif

---

yap that's for sure !^^
What he is saying is that when your room temp goes up 1ºC, your cpu's temp will increase 2ºC. In a rate room-cpu of 1-2

Regards

Let me give you the truth: A 1 degree rise in ambient will result in a 1 degree rise in core temps. Any discrepancy noted is due to error in measurement (either core temp or room temp, you choose).

HMM do you know what I love this guide but I cant change the ambient temps so it is useless. I can alaways change my cooler though !!

Quote:

---

Originally Posted by freecableguy

Let me give you the truth: A 1 degree rise in ambient will result in a 1 degree rise in core temps. Any discrepancy noted is due to error in measurement (either core temp or room temp, you choose).

---

Well, I'm no statistician nor am I an Intel engineer, so I can't speak to experimental errors/instrument errors. What I can tell you is that whatever relationship I measured in this experiment gave a pretty tight data set.

@mukmaster - Right, without running an A/C you can't really change your room temp, but you can understand that if the room temp raises 8 °C, you can expect your CPU load temps to increase roughly by 16 °C.

Quote:

---

Originally Posted by graysky

Well, I'm no statistician nor am I an Intel engineer, so I can't speak to experimental errors/instrument errors. What I can tell you is that whatever relationship I measured in this experiment gave a pretty tight data set.

@mukmaster - Right, without running an A/C you can't really change your room temp, but you can understand that if the room temp raises 8 °C, you can expect your CPU load temps to increase roughly by 16 °C.

---

do you even see the problem here with the cpu temp going up by double the increase of the room temp? :lol:

Quote:

---

Originally Posted by CaNNibaL

yap well done ! now all buy air conditioning :D

---

Yes, definitely!
With the summer approaching we need it cool :)

Quote:

---

Originally Posted by Spawne32

do you even see the problem here with the cpu temp going up by double the increase of the room temp? :lol:

---

Well, the data is the data. I think unless you attach a leaf blower or something outrageous like that to your case, it won't be 1:1. You ever stick your hand in there? It's warmer than room temp given all those heat producing components in an enclosed space. It stands to reason that it would be warmer.

Quote:

---

Originally Posted by Spawne32

do you even see the problem here with the cpu temp going up by double the increase of the room temp? :lol:

---

Thermodynamics>you. :slapass: :fact:

Got any proof to the contrary or is that just (hot air) your opinion?

No proof, yet, but id like some kinda explanation as to why his theory is correct, other then numbers in a nicely colored graph on microsoft excel, because to me this sounds like the same thing as me fighting with my mom in the summer over the a/c, she for some reason thinks that 68 degrees in the house in the winter is hotter then 76 degrees in the summer. Go figure.

Quote:

---

Originally Posted by kingdomwinds

measured with TAT. My case is a centurion 5, which as 80mm intake and 120 exhaust. I use tuniq tower but i think i have bad contact.

So far, i have not had any stability problems but my load temps are 62c-65.

Today, it is not has hot as yesterday. So i am idling 47c right now.

---

Buy one of Cooler Master's 4-in-3 Device Modules and install it so that it's blowing air in towards your CPU cooler. I added one to my Centurion 5 and was impressed enough that I bought another one and installed it in my CM Stacker 810 (in addition to the one that came with the case).

Quote:

---

Originally Posted by Spawne32

No proof, yet, but id like some kinda explanation as to why his theory is correct, other then numbers in a nicely colored graph on microsoft excel

---

It's not a theory, just an interpretation of some data errrr... "numbers in a nicely colored graph" as you put it :P I'd challenge you to do a similar experiment and see if your temps track 1:2 like mine did or 1:1 like you are suggesting before you attack the work others have done based purely on your own speculation with no experimental data of your own to back it up.

Quote:

---

Originally Posted by Spawne32

because to me this sounds like the same thing as me fighting with my mom in the summer over the a/c, she for some reason thinks that 68 degrees in the house in the winter is hotter then 76 degrees in the summer. Go figure.

---

I don't understand this last bit... can you explain how this is similar?

Quote:

---

Originally Posted by graysky

It's not a theory, just an interpretation of some data errrr... "numbers in a nicely colored graph" as you put it :P I'd challenge you to do a similar experiment and see if your temps track 1:2 like mine did or 1:1 like you are suggesting before you attack the work others have done based purely on your own speculation with no experimental data of your own to back it up.

---

Well put! :fact: :clap:

Quote:

---

Originally Posted by graysky

It's not a theory, just an interpretation of some data errrr... "numbers in a nicely colored graph" as you put it :P I'd challenge you to do a similar experiment and see if your temps track 1:2 like mine did or 1:1 like you are suggesting before you attack the work others have done based purely on your own speculation with no experimental data of your own to back it up.



I don't understand this last bit... can you explain how this is similar?

---

LOL u wouldnt understand its a family thing, but ill take you up on your offer, as soon as i have some free time. Right now im really busy with my own website and i just got some parts in from thermaltake that need reviewing :)

hehe so this is what I do.. since I know my room is like 80-100F in the summer.
In the night I put warm clothes on, turn on my a/c either low or high and overclock my rig and I'll see how temps will be, then during the day when the a/c is off windows are open , underclock again.

Updated 15-May with two higher temp points. It was like 85 °F outside today and I left the A/C off. Room got hot so I ran two additional points at about 76 and 77. Anyway, as you can see, it's all linear up to 77.5 °F and I suspect it'll continue that way if it gets hotter.

Quote:

---

Originally Posted by Spawne32

No proof, yet, but id like some kinda explanation as to why his theory is correct, other then numbers in a nicely colored graph on microsoft excel, because to me this sounds like the same thing as me fighting with my mom in the summer over the a/c, she for some reason thinks that 68 degrees in the house in the winter is hotter then 76 degrees in the summer. Go figure.

---

You're mom feels cooler in 76F in summer becuase of our built in cooling system. :D
The evaporation of sweat is enough to keep her comfortable. Everyone is different since what you or I may find as a confortable temp will differ. Thanks.

Quote:

---

Originally Posted by shadow_419

You're mom feels cooler in 76F in summer becuase of our built in cooling system. :D
The evaporation of sweat is enough to keep her comfortable. Everyone is different since what you or I may find as a confortable temp will differ. Thanks.

---

i attribute it more to menapause, but ok? :lol:

Updated 19-May: Fixed a small error and put formulas for both °F and °C (of your room temp) on the graph. The relationship is: for every °F change in room temp, the load TJunction changes by 1-1/2 °C. Well, this is true for my system, YMMV...

i have also found the room temp to cpu temp to be directly proportional as per your graph.

i use a/c in summer.

Quote:

---

Originally Posted by graysky

Updated 19-May: Fixed a small error and put formulas for both °F and °C (of your room temp) on the graph. The relationship is: for every °F change in room temp, the load TJunction changes by 1-1/2 °C. Well, this is true for my system, YMMV...

---

The correct formula is 1F=9/5 C. This is true for everyone, as it is just exact definition. There is no real physics here.

As for difference between ambient and core, it is almost exactly 1:1. The deviation from this rule is so small that you cannot detect this.

Quote:

---

Originally Posted by Cronos

The correct formula is 1F=9/5 C. This is true for everyone, as it is just exact definition. There is no real physics here.

As for difference between ambient and core, it is almost exactly 1:1. The deviation from this rule is so small that you cannot detect this.

---

Yep, in the range of temperatures the average location on earth, the thermal conductivity of copper/aluminum/thermal paste/air will not differ great enough for us to see a big difference.

Well, my point in giving both formulas was to let people see the relationships with the empirically derived data. Yeah, you can go from C to F if you multiply with 9/5 then add 32 :)

all that data says to me is his cpu runs 31-34c over ambient temperature it seems pretty cooralitive to me.

also as temperature rises so does leakage so the graph accelarates a tiny bit as teh temp gets higher

Quote:

---

Originally Posted by brandinb

all that data says to me is his cpu runs 31-34c over ambient temperature it seems pretty cooralitive to me.

also as temperature rises so does leakage so the graph accelarates a tiny bit as teh temp gets higher

---

Well, the slope of the line (0.8302) tells you that for every change 1 °F in room temp will change the CPU temp by 0.83 °C.

Quote:

---

Originally Posted by graysky

Well, the slope of the line (1.4954) tells you that for every change 1 °F in room temp will change the CPU temp by 1.5 °C.

---

Huh? Don't you mean it will change 0.5555 C?

OOps, had them backwards.

ya i would rather just stick with one standard of temp measurement really.

Damn I really wish the US would switch to Metric...Celsius is so much easier than this stupid fahrenheit scale. I made the mistake of getting used to C, so now when I want to express the temperature I have to multiply it by 1.8 and adding 32 :rolleyes:

Quote:

---

Originally Posted by serialk11r

Damn I really wish the US would switch to Metric...Celsius is so much easier than this stupid fahrenheit scale. I made the mistake of getting used to C, so now when I want to express the temperature I have to multiply it by 1.8 and adding 32 :rolleyes:

---

you said it brother :D

Ha, we're a bunch of rebels in the US. After all, the country came out of a bloody revolt. I believe that's why we use the F scale, and drive on the right side of the road, and call soccer, what the rest of the world calls, football :)

Well guys.. I just got my new cooler and temps ambient and weather have been high 80's and very humid. My new cooler with stock fan is still giving hardcore 32c .. idle 38c full load. With SilenX Ixtrema 120 90cfm fan 27-29c idle, 37-38c full load .. I love rev.B

@graysky: Thank you very much for this valuable observation and the effort you put in collecting the data. Would be great if you can take a few more data points in the range of 26-30 C ambient, if possible.

Allow me to represent the data a bit differently, by examining the relationship between the ambient temperature and deltaT.

As was observed, the C/W is not fixed, but rather increases with ambitent. However, one wonders whether the relationship is actually linear. Maybe an inverted S-curve centered around a design point (23 Ambient = 25 enclosure?). I think the inclusion of the low temp observations and the use of CPU core temps rather than deltaT may have masked what is actually going on.

Using the last two observations, one may guess that an increase in ambient of 1C further, may actually result in 1+ increase in deltaT, which means 2C increase in core temps per degree ambient (core0).

And if the behaviour is actually non linear, it may get worse very quickly. The implication being that summer temps may hold some surprises and would require caution.

I'm curios whether the 1:2 ambient:CPU change in temperature ratio is a limitaion of air-cooling per say... or merely a limitation of air-cooling inside a PC box...

I don't know where in the world you live and if you get any extreme weather... but it might be interesting to see what happens when your room temperature hits 35º C or even 40º C (86º F - 95º F)...
I find that some systems have a cooling peak and if you go beyond that point... the system is saturated and overwhelmed, to the point that the internal heat skyrockets out of control...
This would indicate that a 1:2 ratio is directly dependant on the capacity of your PC cooling system...

As a test, perhaps you could take off the side covers of your PC and use a 37cm (15 inch) floor/desk fan to cool your system...
Nobody would want a huge fan like that on a permanent basis... if for no other reason than the sheer noise and dust build-up...
1 test during the midday heat and another in the cool of the night should suffice to see whether the ratio is still maintained...

Also, try using CPU Burn... 1 iteration for each core... which should max-out your CPU heat potential... (100% CPU usage NOT= maximum CPU heat).


:):D:eek:
Pascal

My room temps now in the summer have reached up to 37C. I can assure you that the increase in CPU temp is not linear, but it's not that far from it either. Sucks to not have A/C. So frikin hot in here. I'm lucky I watercool, if not I could never maintain my OC during the summer. Not that winter is very cold here, but at least temps don't go as high as 37C! :eek:

Although now that I think about it, my other PC is air cooled and I still maintain the OC during summer. It's of course not as high as the OC in my watercooled rig, but still 3.2GHz on an E6600 in an air cooled rig with such high room temps is not bad at all IMO. ;)

@mechmech: interesting graphs you've generated. If you do a linear regression line on those data, what is the R^2 value?

Quote:

---

Originally Posted by graysky

So what does this mean and why do you care? Well, using these rules of thumb, if it's currently 70 °F in your room, and your average load core temp is 65 °C, you can expect that to change by roughly 0.8 °C for every single °F your room temp. change. Say your room hits 80 °F. Your load core temp should increase from 65 to 73 °C which may be unacceptable to you and you might want to adjust your o/c accordingly. This is just an approximation based on my system. Your mileage may vary...

---

Do you find, or notice, a similar correlation for CPU idle temps...

The results being true, are a little disturbing... since you're at the mercy of the weather and seasons...
I don't always have a need for seriously OC speed but when I do... then it would be nice to know that it is on-tap... not so if it's a particularly hot day out there.
I considered a little pro-active cooling like A/C assisted aircooling when necessary or even a peltier water-chilling boost (for naturally aspirated water cooling)... but condensation can also become an issue...

I guess it's all par for the course in the quest for a dream/extreme machine...


:):D:eek:
Pascal