*OFFICIAL* M18xR1/R1/18 Owners "Are my CPU/GPU Temps OK Thread" - Ask Questions/Get Help Here

CPU and GPU temps are things that many of us seem to be very concerned about from time to time, so here is a thread where you can ask the question, "Are my temps OK" without having to hunt for the right thread to ask and post the information. I'll do my best in my spare time to update the opening post with a record of reported temps organized by CPU and GPU model.

If you have great temps and want to post that as a point of reference for those with concerns, please feel free to do so. (This will help establish what is within the realm of "normal" temps.)

I'm not sure exactly what the format will end up looking like, so bear with me as I try to figure out what works best. To start off, we will simply index links to posts for specific CPU and GPU configurations to minimize the need to sift through every post on every page to find references to the hardware you want to know about. (I think this might work best.)

Bear in mind that temps can vary between two otherwise identical system based on many factors, no the least of which is ambient temperatures. So, if your temps are slightly higher than someone else's don't freak out about it. For most CPU and GPU the threshold for thermal shutdown is 100°C.

When posting your question (or observations) about your temps, please include the following whenever possible:

• CPU Model (2720QM, 2860QM, 2920XM, 2960XM, etc.)
• GPU Model (6970M, 6990M CrossFire, GTX 580M SLI, etc.)
• GPU Driver (if applicable; i.e. Catalyst 12.5, Verde 296.10, etc.)
• Activity when temps noted (idle, Vantage, Metro 2033, BF3, etc.)
• Ambient/Room Temperature (best estimate if exact temp is unknown)
• Your Thermal Interface Material (stock/factory paste, re-pasted IC Diamond, etc.)
• Optional: Screen Shot Showing Temps (sidebar gadgets, GPU-Z, HWiNFO64, MSI Afterburner, etc.)

​

Remember: Keeping the fans and heat sinks clear of dust and lint is absolutely critical. It takes very little to cause serious heat problems, especially if you use the machine in a warm environment. This problem is the leading cause of overheating and performance issues for CPUs and GPUs. In most cases, replacing thermal paste and pad is not required to correct overheating issues. Using compressed air to blow through the outside vents also does not cut it. It just temporarily moves the stuff and in not time it will be clogging the heat sink radiators again. Pull off the bottom cover, remove the fans and do it the right way. Use compressed air or a vacuum cleaner to get it all of the crud out. Hold the fans to prevent them from spinning up to a high RPM as this can damage them.

Intel Core i7 XM (Extreme CPU)

Mr. Fox - 2920XM - Idle
Quontum - 3920XM - Battlefield 3

Intel Core i7 QM (Non-Extreme CPU)

dave98 - 2760QM - Batman Arkham Asylum
mpearson - 2860QM - 3DMark11, Vantage
Trance003 - 2860QM - Battlefield 3

AMD Blackcomb GPU (6970M/6990M)

UsmanG - 6990M CrossFire - Furmark

AMD Wimbleton GPU (7970M)

Quontum - 7970M CrossFire - Battlefield 3

NVIDIA Fermi GPU (All Models)

mpearson - GTX 560M SLI - 3DMark11, Vantage
Trance003 - GTX 580M SLI - Battlefield 3
dave98 - GTX 580M SLI - Batman Arkham Asylum
Mr. Fox - GTX 580M SLI - Idle

(NOTE: Index is no longer being maintained)

.
.​

 

Hey,

I just ran the 15 min FunMark test and it seems like the temp levels are okay, but what i do not understand is GPU #3. My system only has 2 GPU's right? Which one is the third? It is running so HOT 91+ degrees

I own an AlienWare M18x R1, less than 1 month old.

Here are the screenshot after the 15 min FunMark test.
http://img135.imageshack.us/img135/8202/funmark15mintest.jpg

Does the temp levels seem okay? What is the meaning of GPU #3?

Thanks.

Regards,
Usman

 

Gpu 3 could be the integrated gpu showing up. Either the hd 3000 or 4000 on IB.

 

My initial gut reaction is "no, your temps are fine" and the problem is Furmark. Those temps are not entirely unheard of using Furmark. There's no good reason to run it unless you're trying to test something very unusual that requires GPU stress tests that exceed real-world GPU loads, and then you wouldn't want to run it for 15 minutes. Doing so can cause damage that would not be covered under your warranty. Furmark is unnecessarily harsh.

If you see 91°C from your GPUs running Vantage or 3DMark11, or any games, then you can consider that a problem that needs to be addressed.

Other than what Sam said, no idea why you would see a GPU #3. Unless something is messed up with Furmark. Perhaps you can post a screen shot so we can better understand what you are referring to.

The IGP should not be visible to the BIOS while discrete graphics are active and vice versa. If you look in Device Manager and see the Intel Graphics and your 6990M cards, then there is a setting in your BIOS that is not correct. The M18x does not have on-the-fly switching and that is handled by a switch rather than software. They should never both be visible if the settings are correct... only one or the other.

 

Hi All,

I've never done any benchmarking before so I thought I'd give it ago.
Would appriciate any feedback.

Everything on the M18X is stock and the benchmarks all run with default settings.

CPU = i7-2860QM
GPU = 560M dual SLI

3D Mark 11 = CPU Package Temp 44C-82C
GPU1 Temp = 52C-66C
GPU2 Temp = 48C-64C
Score = 4154

3D Vantage
CPU Package Temp = 44C-88C
GPU1 Temp = 52C-66C
GPU2 Temp = 48C-63C
Score = 17087

Is the CPU Package Temp on the 3D Vantage test OK?

Thanks

MP

 

Mr Fox speaks the truth. Furmark is not really an accurate measure of how your GPUs will perform under real-world load.

I used to get 91° in real world applications, and my cards were overheating bad. If I ran Furmark at that time, they would hit 102° no problem.

At the moment I get up to 85° or so in Furmark, and about 76° in real games, so yeah, your furmark temps are pretty much spot-on.

Those are excellent temps I think. My GPUs get up to 76° under real-world load, so for the 560Ms that should be pretty normal

 

Yep, agree with locust76... normal/good. No worries. You've got overhead to spare with those temps.

 

Temps look awesome. Take care of her!

 

Like Fox said - it is only possible to switch between the 6990M's or the integrated gpu.

In my case i did the FunMark test on the 6990M's. The integrated gpu was off - i have actually never used the integrated gpu.

FunMark gives me unrealistic high temp levels which are 84 is my case - understood Mr. Fox.

I can only enable either the 6990M's or the integrated gpu, you are right. I have only the 6990M's enabled never used the integrated gpu.

I checked device manager under display adapters and only the 6990M's are listed.

Still i do not understand what GPU #3 is...

http://img135.imageshack.us/img135/8202/funmark15mintest.jpg

That is great..



Does anyone have any idea to what the GPU #3 is ?

 

GPU #3 is NOT another gpu - its merely a temperature sensor on the card - if you look at GPU-Z, those readouts are numbered "GPU Temp. #1/2/3"....3 sensors on each card - 3 readings from different points on the same card....for what its worth, I always remember sensor #3 getting considerably hotter than the other two on my previous AMD setup. Nothing to worry about when using a stress test like Furmark, but if you see temps like that whilst gaming, that would require some attention to be paid, as it would be far too hot.

Also, I wouldnt advise continually using Furmark to stress your gpu's - it has been know to burn up cards beaucse it puts a VERY severe load on them - not like any game would put on them.

 

Thanks a lot brother for clearing this mess up. I won't be using FunMark anymore.
What program should i use to test my gpu then? Something that is more realistic than FunMark and also not damaging to the gpu's.

Another question off topic:
I use prime95 for cpu and ram testing, is this damaging as well?

 

Well from those replys I guess the 88C on the CPC isn't anything to be concerned about.

Thanks for the peace of mind chaps.

MP

 

Cpu's tend to be alot more resilient than gpu's and have built in protection to avoid damage.... but saying that, I would never over stress ANY component more than I have to - as the saying goes "everything is OK, but in moderation" - you will be fine to use prime95 for cpu testing but I wouldnt be constantly benching your cpu.

As for an alternative to Furmark, you could use MSI Kombustor - same kinda program to Furmark but whenever I have called Dell Technical Support, that's the program they have used when diagnosing gpu issues.....again, overuse of it would not be advisable, as with any serious stress testing - it's ok for diagnostic purposes from time to time but constant use could burn up your gpu's....I would say only use either when you really need to diagnose potential issues. They arent like running vantage or mark '11 because they really do place a huge load on your video cards.

 

The best way to test your system is to play games and see how things perform. Metro 2033, Crysis, Crysis 2, Batman Arkham Asylum, Batman Arkham City, STALKER Call of Pripyat, Resident Evil 5, and Just Cause 2 are just a few of the games out there that have a benchmark option. You can also run Futuremark's 3DMark11 and Vantage tests, which are roughly the equivalent of very taxing games. The 3DMark11 and Vantage CPU tests are significantly more stressful than ordinary games, but not to the extent that you would need to be worried about the stress level. If you are having an overheating problem with CPU or GPU that needs to be corrected, you will find it with 3DMark11 and Vantage without breaking anything in the process.

 

Peace of mind is priceless. Especially so on such an expensive machine.

 

I used to have a mobile i7 which got to that temp, its normal. But if you really want to see what your temp will go to, run a 3hr or so torture test on prime95 on the small test (most stressing)

 

I have only run FunMark 2-3 times total and the same goes for Prime95. My M18X is only about a month old - I read many people have had problems with their M18X, so i just wanted to confirm that my system is problem free. I ran the tests only to see if i encountered any problem when overloading the CPU and GPU's - I guess it's problem free I am a happy M18X owner!

Thanks a lot StevieJones and Mr. Fox..

Your help is highly appreciated..

Thanks again!

May god bless you.

Regards,
Usman

 

No problem UsmanG, good thing about NBR is we all like to help each other when we can......

 

Why?????????????? - unless you want to encourage problems.

It's somewhat pointless to put your system under undue stress for fun....unless you have more money than sense.

 

Thanks. I was about to respond in a similar manner. There's simply no point in doing that. I'd say a 3 to 5 minute stress test is more than ample time for an overheating problem to become apparent.

Conceptually, that's about like, "Hey, let's throw a cinder block on the accelerator pedal of the car to see if the engine blows up. If it doesn't, we're good to go."

Stress testing has a very limited-scope value in a scenario where one needs to confirm that an existing problem has been successfully corrected. Torture testing creates new problems that typically could (and should) be considered owner-induced damage that is not covered under warranty.

 

Hi all. I've been using my M18x for about 5 1/2 months now and it's absolutely fantastic! However, the temps of the processor is really worrying me. First off, i'm using CPUID monitor. I was playing battlefield 3 on AUTO settings for about approx. 1 1/2 hours and the temperatures i got are the following (in Celcius):

GPU 1:
Min - 42
Max - 69

GPU 2:
Min - 45
Max - 70

Processor (2860QM):
Core 0 - 43 Min, 89 Max
Core 1 - 42 Min, 93 Max
Core 2 - 42 Min, 86 Max
Core 3 - 41 Min, 86 Max
Package - 43 Min, 93 Max

The figures shown are really bothering me! The only concern here is the processor. Is it normal? I even use a notebook cooler while gaming and my CPU+GPUs are coated with IC diamond 7 thermal compound. All of these don't really add up. Is there anything wrong? I do appreciate your comments and advise for this problem. Thanks for your time.
P/S: Ignore the other two 'Temperature concerns' threads. My internet had some lags and I think I clicked on submit thread too much. =P

 

I would check for dust, clean the vents in the back. Might not hurt to take the back cover off to clean them.

 

I'm happy that I have piece of mind. Thats all I needed.
When you have a machine that costs this amount of money you want to look after it. Any sign of trouble and its time to check in with the users in this forum and then on the phone to Dell.

As mentioned in one of my other threads she is about to get an upgrade with a SSD and a 1TB storage drive.

Should be fun. Hope she copes with it all.

Crucial M4 I think Mr Fox.

Cheers MP

 

You don't have to butcher the laptop to make issues visible. Just run a game that you know is intensive on other systems. Keep a close eye and see what happens.

The GPU's look normal to me, if they're around 70* on heavy load then that seems acceptable.

I could only suggest using a can of compressed air on the fans, I wouldnt really want to tear the lappy apart to clean it, but depending your expertise its upto you.

 

It takes four screws to take the back cover off and then you can see the cooling fins to make sure they are clean. Using compressed air from the back will just throw the dust back into the computer, and once the fans start up again, all the dust will be back on the fins.

 

Your gpu temps look perfectly normal to me. Your cpu is getting a bit too hot for comfort. If after cleaning the system fans (and heatsinks too would be a good idea) you are still getting high cpu temps, it might be because of your TIM - noticed you repasted your chips and might be the case that it needs doing again for the cpu.....aftermarket compounds tend to require re-application more frequently vs. Dell stock compound....the aftermarket stuff may perform slightly better in the short term but then temps can creep up relatively quickly after its been applied.

 

Dang, another new thread with questions on temps? I think I've lost count, LOL. Maybe we should start a master thread for questions on temps by combining all of them.

 

I played Battlefield 3 for like 5 minutes with the settings on ultra, vsync off, and RealTemp was reading max temp of 77C on the CPU, is that high or normal? Seems high to me but I'm used to seeing high max temp around the 60s on my desktop machine.

 

Seems perfectly normal to me......

 

Yes, I agree. No reason to be concerned about 77°C, as it's not too hot by any means. The only folks that would need to fret about that are owner's of an M17x R3 with the 78°C throttle design defect.

 

I just typed a long reply so here is the cliffs notes.
1. Great idea on a sticky
2. People, please post your specs before asking about temps. Maybe add that to the original post.
3. Maybe create a rough reference of GPU temps for known CPUs and GPUs.

Obviously people will still post temps because the differing components will impact the total heat. A 2960xm and 2 750HDDs will make a 580m hotter than a 580m with a single HDD and a 2720 (all things being equal).

Here are my temps (rig in sig):
Game: Batman AA
CPU Temp: 75-81
GPU1 Temp: 77
GPU2 Temp: 73

I NEVER use furmark or other stability testing. Though my clocks are stock, if they were not, then I would still use gaming as my "stability test."

 

Thanks for the suggestions, Brother Dave. Good ideas.

And I think your temps look great, buddy.

How does this format look for the opening post?

 

Here are my normal idle temps:

2920XM at 4.2GHz - 64°C
580M SLI (stock, 0.92v)
GPU0: 42°-47°C / GPU1: 42°-47°C

 

Around what temps should I be worried about for 3920XM, I have it on Overclock Level 3 and one of my cores max temp was at 88C, though its probably still good because the TJunction is at 105C, but I just feel like the fans never really spin up.

Messing around with Intel Xtreme Tuning Utility I just hit 90C on all cores.

 

Is 64°C not really hot?

 

64c is perfectly within the normal range. Just fine

 

Is 94 degree too hot for running wprime 1024M. trying to see how far I can push the 3920xm

thanks

 

Depends on how much you have overclocked the cpu to get 94c.....if you havent OC'd much, its too hot...if you have a heavy overclock, it aint soooo bad.

 

43 on all 4 cores. I manage to lower it to 93 playing with bio settings. thanks for your reply.

 

After 1/2 Hour of Metro 2033. Is this okay durring game play ? I think my CPU temps are running a little high. I'm also a little paranoid now Mr Fox enlightened me on the fact I should have a 3 pipe Heat sink instead of two pipe.

CPU Model: 2960XM
GPU Model: 6990M CrossFireX
GPU DriverCatalyst: 12.5
Activity when temps noted: Metro 2033, Full RES Max Settings with DirectX11 and Enhanced PhysX.



30 mins after letting system idle:

 

Looks good to me, Brother Matrix!

 

No screenshot here, but according to HWMonitor, my temps don't go above 75 C when I play games. Is this good enough?

 

If your using the specs you have in your signature I would say yes. I ran those specs for a year with never pushing 80 C. I also Overclocked the 460M and never pushed much higher either and had about a 30% performance increase.

 

I agree... I think your temps are good Vergil.

 

I have had this laptop now for 3 weeks and I have had 2 random games freeze on me where I couldn't even Ctrl,Alt,Del to get to my desktop and had to do a hard reboot.
The game I was playing was BF3 at full ultra settings with FPS at 54-58.
Game time when the screen froze was roughly 90 to 120 minutes of game time.
The other game was SWTOR at max settings, game time playing was 3hrs at FPS of 109 before screen froze.
Any idea what my problems are? again, they are random and only happen twice. Room temps playing were 65 F.
I had a laptop cooler which is considered the best.

Thanks for reading.



+- Alienware M18xR2 (/mainboard)
|
+- Intel Core i7-3820QM (/intelcpu/0)
| +- Bus Speed : 448.992 448.959 449.024 (/intelcpu/0/clock/0)
| +- CPU Core #1 : 5387.9 1346.88 5388.29 (/intelcpu/0/clock/1)
| +- CPU Core #2 : 5387.9 1346.88 5388.29 (/intelcpu/0/clock/2)
| +- CPU Core #3 : 5387.9 1346.88 5388 (/intelcpu/0/clock/3)
| +- CPU Core #4 : 5387.9 1346.88 5388 (/intelcpu/0/clock/4)
| +- CPU Core #1 : 46 37 70 (/intelcpu/0/temperature/0)
| +- CPU Core #2 : 46 37 71 (/intelcpu/0/temperature/1)
| +- CPU Core #3 : 46 36 71 (/intelcpu/0/temperature/2)
| +- CPU Core #4 : 46 33 69 (/intelcpu/0/temperature/3)
| +- CPU Package : 46 38 71 (/intelcpu/0/temperature/4)
| +- CPU Total : 0.192308 0 77.6923 (/intelcpu/0/load/0)
| +- CPU Core #1 : 0 0 83.0769 (/intelcpu/0/load/1)
| +- CPU Core #2 : 0 0 69.2308 (/intelcpu/0/load/2)
| +- CPU Core #3 : 0 0 75.3846 (/intelcpu/0/load/3)
| +- CPU Core #4 : 0.769234 0 86.1539 (/intelcpu/0/load/4)
|
+- NVIDIA GeForce GTX 675M (/nvidiagpu/0)
| +- GPU Core : 405 0 620.156 (/nvidiagpu/0/clock/0)
| +- GPU Memory : 1494.82 135 1494.82 (/nvidiagpu/0/clock/1)
| +- GPU Shader : 810 0 1240.31 (/nvidiagpu/0/clock/2)
| +- GPU Core : 62 43 71 (/nvidiagpu/0/temperature/0)
| +- GPU Core : 0 0 99 (/nvidiagpu/0/load/0)
| +- GPU Memory Controller : 0 0 51 (/nvidiagpu/0/load/1)
| +- GPU Video Engine : 0 0 0 (/nvidiagpu/0/load/2)
| +- GPU Memory : 7.901 6.47717 70.3434 (/nvidiagpu/0/load/3)
| +- GPU Fan : 100 100 100 (/nvidiagpu/0/control/0)
|
+- NVIDIA GeForce GTX 675M (/nvidiagpu/1)
| +- GPU Core : 405 0 620.156 (/nvidiagpu/1/clock/0)
| +- GPU Memory : 1494.82 135 1494.82 (/nvidiagpu/1/clock/1)
| +- GPU Shader : 810 0 1240.31 (/nvidiagpu/1/clock/2)
| +- GPU Core : 63 46 77 (/nvidiagpu/1/temperature/0)
| +- GPU Core : 0 0 98 (/nvidiagpu/1/load/0)
| +- GPU Memory Controller : 0 0 51 (/nvidiagpu/1/load/1)
| +- GPU Video Engine : 0 0 0 (/nvidiagpu/1/load/2)
| +- GPU Memory : 7.901 6.47717 70.3434 (/nvidiagpu/1/load/3)
|
+- ST9500423AS (/hdd/0)
| +- Temperature : 38 34 40 (/hdd/0/temperature/0)
|
+- SAMSUNG SSD PM830 mSATA 64GB (/hdd/1)
| +- Remaining Life : 92 92 92 (/hdd/1/level/0)

--------------------------------------------------------------------------------

Parameters

|
+- Alienware M18xR2 (/mainboard)
|
+- Intel Core i7-3820QM (/intelcpu/0)
| |
| +- CPU Core #1 (/intelcpu/0/temperature/0)
| | +- TjMax [°C] : 100 : 100
| | +- TSlope [°C] : 1 : 1
| |
| +- CPU Core #2 (/intelcpu/0/temperature/1)
| | +- TjMax [°C] : 100 : 100
| | +- TSlope [°C] : 1 : 1
| |
| +- CPU Core #3 (/intelcpu/0/temperature/2)
| | +- TjMax [°C] : 100 : 100
| | +- TSlope [°C] : 1 : 1
| |
| +- CPU Core #4 (/intelcpu/0/temperature/3)
| | +- TjMax [°C] : 100 : 100
| | +- TSlope [°C] : 1 : 1
| |
| +- CPU Package (/intelcpu/0/temperature/4)
| | +- TjMax [°C] : 100 : 100
| | +- TSlope [°C] : 1 : 1
|
+- NVIDIA GeForce GTX 675M (/nvidiagpu/0)
|
+- NVIDIA GeForce GTX 675M (/nvidiagpu/1)
|
+- ST9500423AS (/hdd/0)
|
+- SAMSUNG SSD PM830 mSATA 64GB (/hdd/1)







M18xR2
Intel Core i7-3820QM
Dual 2GB NVIDIA GeForce GTX 675M - SLI - Driver Ver: 8.17.12.9639
16GB Dual Channel DDR3 at 1600MHz
SAMSUNG SSD PM830 mSATA 64GB

 

guys is it good to to do a restore on a brand new m18x with restore cd's it came with? i already have games installed...

sny benefits from doing it and what will i lose?

 

M18x R2 @ idle with an ambient temp of about 75 F
i7 3720QM @ 2.60 @29c
7970m xfire with gpu1 @ 54c gpu2 @ 59c

Are these temps normal for idle? I have upgraded to the diamond 24ct but these gpu temps still seem high for idle. Thanks for any input.

 

If I understand your question correctly, you will lose everything (user generated files) if you Respawn to a factory state. If you do a partial restore you should not lose all files, but the installation information in the registry will be lost. If there is something wrong with how your system is working, you may have to do that. If your system is working properly, create an image so you can restore it to the current condition with everything installed the way you want it.

This is a driver problem or an overclocking problem. If you are not overclocked, it is probably a driver TDR issue. AMD 7970M has the same problem, but it seems your system with 675M SLI is not having it as frequently as 7970M CF.

Those idle temps are just fine. No reason for concern at all.

 

Well with temp seems like every system have their own rules mine Gpus ideal temp. are Gpu1 54C while Gpu2 44C. so its like completely opp. then yours.

 

Great, wasnt sure if i did an improper job with the thermal compound. Glad to hear its atleast normal though i would have hoped for some drop with the IC diamond application.