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

i've heard that the U3 with or without fans is the best choice, ive heard people have had increased temps with the CF 19 coolermaster one

Also i've changed to 40/39/38/38 without enabling overclocking level 1
and i dont see a change in, still remains at 3.8

 

The CF-19 is not the best for the m18x but its better then nothing and about the same as raising the back of the laptop temperature wise.

But I had it before I got my m18x ...

edit: try with overclocking at lvl1 it might help
but I'm thinking you just dont see it happen because each core will not stay long at 4.0 since the load is always changing core and monitoring apps only refresh like once or twice per second

 

i'd imagine level 1 overclock just increases voltage as with 2 and 3, confirm/deny?

 

It is possible that 3.8GHz is a fused clock speed limit with your CPU. If that's true, it might ignore the multiplier settings that are higher than the fused limit. If you can manually change the individual core multipliers, try setting all 4 at 40 and see what happens. The worst case scenario would be you will still be maxed out at 3.8GHz.

You should be able to see the flex (voltage) setting change when you switch between Level 1, 2 and 3. With a locked BIOS, there could be some other changes taking place that you cannot see with hidden menus.

With the unlocked BIOS you can disable the MSR lock, and that should allow (at least with an XM CPU it should) higher clock speeds.

 

I have seen my 3720qm hit 4.0 (with cpu-z and one instance/thread of prime95 (right click on the cpuz window to see clock for all 4 core))

 

What your temps during prime95? (3piped heatsink?)

 

So, was it a brief 4GHz spike or a sustained 4GHz? What is the highest clock speed you have sustained running on all 4 cores?

I only ask because I have seen a very brief spike a few times as high as 9GHz on one core with the 2920XM cranked up to 4.8GHz, but it lasts less than 1 second and really is meaningless.

 

I went back into the bios and just resaved my settings and now i am seeing brief 3.96 or something, so i assume its working as intended now

 

CPU temp during Prime95?

 

Here is what I've got for a ~11 min run of prime95 with only one thread (to get to 4.0ghz which is highlighted in the cpuz windows in the screenshot) :

as you can see in the screenshot the cpu load (green line) is always changing core so each core get up to 4.0ghz in a rotation (I suppose its to keep it as cool as possible) and I'm thinking that sometimes the monitoring apps will not refresh fast enough to always show it right.

from a previous post in this thread :

The max I can sustain on 4 core is 3.8ghz. Since most game are not supporting more then 2 threads its a way to get an extra 100/200mhz more

*I have the 2 pipe heatsink.

 

Have you tried bumping up your BCLK yet? That will push it higher still and will help with speed of some other components as well. Try 102.5Mhz to start with and see how it does, and adjust up or down from there.

The fluctuation might be something built into the CPU that is causing it to throttle. If you have flashed the unlocked BIOS, go to the CPU Performance menu and disable MSR lock. If it will let you disable c-states, do that also. This should stop the fluctuating clock speeds. At least it does with Sandy Bridge XM CPU.

 

I have not tried to bump up the BCLK and I think I'm gonna pass on this one. It could be nice for benching but I would not do it for normal use.

I believe that what I described about the cpu load changing core is normal behavior. My previous desktop with a 2500k was showing the exact same behavior and the same thing with my AMD 1090T setup before that (its even easier to spot with a 6 core cpu). If I'm not mistaken they do it this way for power and thermal reason. Anyway if the core is not actively working it does not need to be clocked all the way up to max turbo speed and thus using less power and generating less heat.

 

Xmadror, what your ambient temps? Mine is 30 celsius now

 

These days, I'm usually at 24-25C (or lower).

 

Well, 30°C (86°F) ambient temperatures is a pretty hot environment for any sort of overclocking. What are ambient temps of 30°C producing at idle and under load for your CPU core temps running at stock? Ideal ambient temperatures would be less than 22°C (70°F). Are you not working in an air conditioned environment?

 

My stock temps at idle: CPU hovering ~50 celsius, GPU ~50 celsius, when it was cooler it was the same... i think with 2 piped heatsink which i bend a bit i was getting 88Celsis during prime95 (30mins) now im getting 92 celsius with 3 piped, something is wrong here..

 

Maybe not a problem, bro... what is your CPU clocked at? Running 30 minutes of prime95 might be way more stress than you would ever see playing games or benching. What do your temps max out at running Vantage? If you don't end up with a thermal shutdown running both Vantage CPU tests, you may never experience a serious issue with CPU temps during normal use.

 

During dirt showdown its 85 Celsius, 3Dmark11 82-83C, now i'm playing Sleeping dogs and getting 89C. Is that normal? I have 400MHZ overclock on my 3820QM

 

You're probably OK. How about Vantage? If you can pass both CPU tests without shutting down, you're probably going to be fine. It's pretty warm where you are, so that might be as good as you'll get running with an OC.

 

Didn't bought vantage yet... i will try to repaste with a line of ic diamond instead of 2 pea blobs... also i will try to bend heatsink like i did with 2 piped.

 

Its totally possible that +5 or 6C ambient temp lead to an increase of more the 5-6C on intensive task.
The hotter the ambient temps are the less effective its gonna be at cooling down that heatsink.

Its weird that you had better temps with that 2 pipe of yours. At 88C (with 2 pipes) you were ~8C higher then me while having an ambient temp of 5-6C higher.
That seems plausible to me.

But 92C with the 3 pipe, I think something is wrong.
Is that with the dell "stock" thermal paste or IC Diamond ?

I have not done any real testing regarding the stock Dell paste but judging from what was left on my 2 pipe and 1 pipe the stock thermal paste application from Dell is pretty thick and it could lead to increased temp just because there is too much of the stuff.
But that's just speculation because I went from the 1 pipe with stock dell paste to the 2 pipe with AC MX-2. So I have not compared result with just changing the paste.

 

I have removed dell's stock paste and applied IC Diamond, so should i leave things as is or try to repaste? You have used a line method when repasting?

 

I am finding playing Source based games: Counter-Strike Global Offensive Beta I am getting 79-84 on my i7 2760qm. Is this fine? I know it has a thermal max of 100 before throttling but is running 84 safe for long gaming periods? My GPU's have been replaced and repasted and am running a cool 69-71 on both cards. Thanks for the input in advance.

 

84C is ok for long run i think.

 

I'd try a repaste. After a repaste on my 2 pipe heatsink I ended up 3-4C lower then the 1st try so maybe it'll help.

And yes I've used the line method with a very thin line in the middle covering ~75% of the length of the cpu die.

 

Would line method work aswell with IC diamond?

 

^^ I can't see why not

 

repasted with line of IC Diamond, exactly the same picture lol 92 - 93 celsius during prime95

 

Still stressing on this prime95 thing? I get the same temps with prime95, bro. It doesn't matter because I don't see those temps when doing normal computing. If you don't see them except for doing extremely intensive CPU stress tests, it will not affect your overall experience.

Here's a link to download for the free version of Vantage. You don't need to buy it. Download 3DMark Vantage | 3DMark 11 – The Gamer's Benchmark for DirectX 11

 

88Celsius after successfull vantage run. Is that ok?

 

Yes, if that's the peak temp in Vantage you're in good shape, bro.

 

yeah, its peak. Interesting thing is Sleeping dogs game is stressing my cpu as vantage?! I am getting 88 Celsius during gameplay.

 

That's really not a problem as I see it. With your high ambient temps, this doesn't seem all that bad to me... maybe not ideal, but it's already pretty warm in the working environment. It sounds like the game is using your CPU and 88°C is not extremely/dangerously hot. Do other games get the CPU warmed up a bit, or just this one?

 

just this game... dirt showdown gives max of 83 Celsius.... Just cause 2 max 80Celsius

 

Well then, that's really good news.

 

Try this game it has a good scenario and also you could report ur cpu temps aswell

 

I would be happy to, but will not be able to at $49.95 USD. That's about double what I am willing to pay for games. I almost never pay more than $19.95 even for a game that I am dying to get my hands on. Maybe when the price drops that low and I can get it on a good Steam weekend sale.

 

Ok, are you getting high cpu temps (85-90c) in any games you played?

 

Mine are very similar to yours, but you cannot really compare our operating conditions very well. I am running 4.5GHz on all four cores and my air conditioned house is almost always 68°F (20°C) or cooler. It's only 85°F (29°C) outside right now where I live. The ambient temperatures make a huge difference. If I was in a similarly warm temperature zone, I would have to lower my CPU clock speeds to avoid heat problems.

 

Just installed my triple pipe cooler and recieved my notepal u3, and with both, i'm over 10c drop!! fantastic cooler! and using stock dell paste, impressive thus far, cpu fan rarely kicks in and max temps during battlefield 3 gaming is 75C and usually never above 69-70, and thats in a warm enviroment, with AC on, nothing above 66C !!
thats overclocked as well don't remember exact settings but 3.8ghz and random spikes to 4.1ghz during the short phase, which i'd love to just have 4.0 or 4.1 on all cores, how could i achieve this?.

Many thanks to Mr Fox and his Tear down video, helped a ton.

 

Whats ur prime95 temps? (run it for 30mins)

IS 83 Celsius ok after finishing 3Dmark11?

 

I took delivery of my Alienware M18x R2. I just want to know if these temperatures are okay.
Specs: CPU = i7-3920XM with OC up to 4.1 GHz. GPU = DELL GTX 680M SLI

Idle
CPU: 48-54 deg C
GPU: 39/44 deg C

After playing Batman: Arkham City for about 45 minutes
CPU Max (from HwInfo): 81 deg C
GPU#0 Max: 65 deg C
GPU#1 Max: 74 deg C

Most of the time I could probably tell from EVGA Precision w/ HwINFO's info also that the averages were APPROX
CPU Average Approximation: 66-68 deg C
GPU#0 Average Approximation: 57-58 deg C
GPU#1 Average Approximation: 68-70 deg C

Playing on all settings max (except anti-aliasing which was on 16x QCSAA instead of 32x CSAA (I don't really see a point going any higher than 32x and 16xQ has more colour samples), also DX11 mode of course.

 

Dual GPU nVidia 675M playing F1 2011 high resolution
temp is 75 - 81 C
is ok or not ?

 

680m in sli, is it normal for GPU2 to be around 5-10c hotter then GPU1? aka one GPU at around 45 the other around 54ish... i know the temps aren't bad, they dont get very hot under load in most gaming i dont see GPU 2 over 70, but its still 5-10ish hotter, is that because GPU two is next to the CPU and the additional heat is transferring?

so after running furmark at 1920x1080 8x AA for 10 minutes, they finally finally stabilized at GPU 1 60c GPU 2 70C and with Prime95 running aswell GPU 2 went up to 72C, could this be bad TIM, Bad Contact with heatsink? or just the naturally extra heat in that area from CPU, if i remember correctly GPU to is to the Right side of keyboard.

 

^^ generally speaking my 7970m are about the same temp when under load, at idle one is always 5 to 8C hotter.

 

unless its the new drivers doing this... upgraded to 306.02, but still one shouldnt getting hotter then other, clock and gpu load are the same within small margins, wouldn't make a 10c difference.

 

Outstanding results, Annie. You're in great shape with the temps. CPU and GPU temps are fine. Enjoy your beast.

This is normal guys... totally fine. All dual GPU systems have this kind of variance in temps and the maximum temp you are seeing is well within the safe range. No need for worries about it. The left GPU is centered on the fan and the right GPU is slightly offset. That reflects the variance in idle temps. Plus, differences in load while gaming can also raise temps on the card that is working slightly harder than the other.

Running Furmark for 10 minutes? You might regret doing that. Unless used in moderation and very short bursts, Furmark is known to have fried AMD and NVIDIA GPUs. This is a brutal thing to do to your video cards and I would avoid it... just my suggestion.

 

so its normal to see 5-10c difference in load? my idle temps are the same roughly just load it gets hotter, GPU2 that is.

 

Yes, as I mentioned, it's totally normal. If you look at the system while it is apart, the left GPU is centered on the fan. The right GPU is not. On the GTX 580M they actually used a completely different support plate on the back of the GPU on the right side to help even out the temps between them. At any rate, it is no reason for concern. And, the temperatures are nowhere near high enough to be a problem.

There can be some degree of variance between systems, too. As long as the temps are in the reasonable range, you're good to go.

 

Hi,

I have three week old M18X R2 with:

- dual i7-3920XM running at 3.1Ghz
- dual Nvidia GeForce GTX 675M running in SLI mode
- Samsung 500Gb SSD
- 1Tb 5200rpm HD
- 32Gb RAM
- Killer Wireless

After 1/2 an hour of BF3 with everything turned to Ultra I get GPU 1 & 2 @ 71C, Core 0/1/2/3 all at around 59C

The fans seem to be blowing fairly hard and there is a lot of heat coming out of the back. The notebook is sat on a hard top desk with air able to get to it. I'm just noticing that SpeedFan 4 has little flames next to the GPU and Cores, I'm not concerned. Should I be?

Thanks,

Matt.