M11xR3 Full load CPU/GPU Temp Thread

Folks,

Can you please post your m11x r3 only full load CPU and GPU temps please (Highest you see them go)

Trying to get a handle on what normal temps should be for this computer - spoke to dell rep and he said between 50-70 deg C, yet we seem to have a few people with GPU/CPU's that go significantly higher....

Before I reapplied thermal paste mine were cpu/GPU - 88/96

After I now get 84/84

Still seem high to me....

Thanks,

 

I would get people to run something specific and use the same program as you to get the temps. That way you will get more accurate results.

 

I was playing MoH with some buds for over an hour and when I backed out it was mid 80s across the board. CPU gpu all were hovering around 85c. I think I will do a tear down and replace thermal with diamond IC 24kt and check again. When using a cooling pad the highest I can remember is like high 70s

 

thanks for the info! Hopefully some more people will let us know and we can see what the best are getting and gives us something to aim for - don't want to bother trying to get dell to replace my board if 85's is realistically the best we are going to get with this higher TDP GPU. (35 W I think vs the GT335's 28W)

 

I've been getting very low temps on benchmarks and playing games (mainly COD: Black Ops and some Fallout 3 since I've had the system). Nothing over about 55 C so far, either on the GPU or CPU. I use CPUID and Speedfan for measurements. I haven't played more than 2 or so hours straight though.

If there are specific programs or games you'd like us to test out, let me know.

 

Wow cool

 

I'm assuming that's not with a note book cooler ? How about running intel burn test or occt to test CPU/gpu at full load. Those are awesome temps for a r3... Time to talk to Dell maybe...

 

No laptop cooler, it's a bit chilly in my room usually though. Someone posted similar temps under load here but can't find the thread. I'll dl those tools and get back w/ more readings.

 

the people who give out their temps.. can you also write what country your in.... that makes a HUGE difference.. i travel alot and notice a big difference in the temps.. Im in brasil right now and i wouldn't even think about playing a game for more then a hour.. I go to london england on friday. the temps there are completely different.....

 

How many times should I let the IntelBurn Test cycle through? It's on Maximum, the rest were defaults. @MarkTrav82, I'm in Portland, Oregon.

1st test:

http://imageshack.us/photo/my-images/839/intelspeedtest1.png/

2nd test:

http://imageshack.us/photo/my-images/269/intelspeedtest2.png/

3rd test:

http://imageshack.us/photo/my-images/709/intelspeedtest3.png/

4th test:


http://imageshack.us/photo/my-images/192/intelspeedtest4.png/

5th test:

http://imageshack.us/photo/my-images/218/intelspeedtest5.png/

Not sure if that was helpful; didn't get a good screengrab of the first test and may have waited a few min too long to take the final shot. Reading through forums, it seems people run that test anywhere from 5 to hundreds of times. The CPU temps didn't go over 67 C, for what's it's worth. It definitely made my system warmer that plain old gaming does. I'll try some more tests tomorrow.

 

Note the programs used to measure the temperature- Speedfan is notorious for underreporting temperatures.

No it doesn't. Location doesn't help at all. You must mean ambient temps instead.

 

I didn't use Speedfan for this last test, but HWinfo32 sensors. Is that one at all reputable? Do you suggest others?

I'd say location certainly effects ambient temps. I used to live in the Mojave Desert, CA without AC. Wouldn't be gaming long in summer in that heat. ;-)

 

I don't use HWinfo32 so I don';t know. I suggest Core Temp, TealTemp or Throttlestop for CPU temps and GPUZ for the GT540M.

Location certainly affects ambient temps but it doesn't account for room conditions- it simply presumes the user is in outside in that climate, not accounting for heating/cooling of an environment. How valid is saying a user is in the Arctic so he must be having sub zero temperatures? He or she could be in a 30C heated room for all we know. Plus, it doesn't help users unaware of the location's climate- what is Oregon's climate to an Australian?
Reporting ambient temps is much better then location as it's much easier to reproduce somewhere else in the world and is an actual variable instead of location.

 

Thanks for the suggestions! Reporting ambient temp definitely makes sense for accurate readings.

It's funny, overclocking and testing all this stuff is almost as fun as gaming. =)

 

If you want to compare temps, do this: (And do it exactly. Don't skip steps or cut the times short or anything....EXACTLY like this)



Step 1: Download Prime95, RealTemp, & ATITool.
Step 2: Sit your machine on a HARD, FLAT surface. Let the machine sit for 15 minutes. (don't use any notebook coolers.....don't sit it next to an air conditioner, just a normal hard flat surface in a normal room hopefully close to room temperature)
Step 3: Launch RealTemp. Write down your idle temps. Also write down the temperature of the room the machine is in (This will be your ambient temperature. Very important)
Step 4: Run Prime95, use in-place Large FFTs. Run it for 1 hour. (It can take several hours to reach your maximum temp.....but for comparison's sake....everybody run it or 1 hour) Write down your maximum temps measured by RealTemp. These 2 (or 4) temps will be your CPU load temps.
Step 5: Turn your machine off and let it cool down for an hour.
Step 6: Turn your machine back on & make sure you disable your screensaver. Also make sure your monitor doesn't shut off after a time period.
Step 7: Launch ATITool. Check your GPU's temperature right away. Write this down. This will be your GPU idle temperature.
Step 8: Turn on the 3D view (If it's not already)
Step 9: Let it sit like that for an hour.
Step 10: Check the GPU temp after the hour is up. Write it down. This will be your GPU load temperature.




Report back here with:

Room temperature
Idle Temperatures(CPU, each core)
Idle Temperature(GPU)
Load Temperatures(CPU, each core)
Load Temperature(GPU)



Compare & contrast. Take note that the temperature of the room will affect the other 4 temperatures. Since this is air cooling, the ambient temps of the room play a big part.



This is the method I use to check temps when I build a PC(Although I run Prime95 for 12-24 hours). It may have to be adjusted due to that Optimus stuff....but it should work fine. (I only have a R1 model, so I don't know how optimus affects this) If ATITool gives you trouble, you can substitute in EVGA OC scanner instead. Tho, the process will be a little bit different (You'll have to set a resolution, and you'll have to use the artifact scanner.....it wont raise the temps to max, but if everyone does the same thing you can compare temps that way) The key here is to get everyone to do the SAME EXACT METHOD OF TESTING. Also, to make note of the room temperature when the testing occurred. Just saying "HURR DURR I GOTZ 85 DEGREEZ IN TEH CRYSIS TOO" means nothing as different parts of different games stress the CPU & GPU differently....even if you play through the same game your playstyle can vary which can vary your temps from another user's. Also notice that my method tests the CPU & GPU individually, so you wont get heat-soak from the other component while trying to test just one.



Gentlemen, lets see those temps.

 

RealTemp 3.67
RealTemp_367.zip

Includes Sandy Bridge VID and power consumption reporting.