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

IC Diamond may work better after a day or two, once it has fully melted and spread out.

 

guys, i just recived my m18x r2 with dual 7970s i noticed after some time of just web browsing one of my gpu temps show 0c under speedfan.

is this normal????? its almost like one of the gpus goes to sleep. that didnt used to happen with my dual 6990s

 

Probably power saving state on the secondary card.

 

Hopefully someone with an R2 can download speedfan 4.46 and confirm this... just got this baby and wanna make sure its working as it should.

 

I have SpeedFan 4.46 installed. No 7970's but I have 6990's and the second listing for the GPU always shows Blue Arrow down with 0c. Keep SpeedFan running in the background and play a graphic intensive game, exit out after 10-15 mins or so and you will probably see your GPU running a working temp. If not I would guess Crossfire is not enabled, your playing a game that isn't taking advantage of both cards or possibly the card is faulty. I wouldn't jump to conclusions just yet though Let me know how it works out.

My Current Screenshot while at 0c



I opened 3dMark11 and saw a temp jump then immediate drop to 0 again at idle.
Also After gaming I can see steady temps and feel both fans.

 

Juang1985 - I think there are issues with the sensors on 7970M. Whether that is a hardware thing or an AMD driver bug, I am not sure.

I don't use SpeedFan because I have had issues with it, including DPC latency and the fans stopping. I use HWiNFO64 instead and it has sensors that show in the system tray. The HWiNFO64 system tray sensors have never stopped working. However, I do have to take my cards out of CrossFire to set the sensors (or they become hidden) and then re-engage CrossFire once they are set.

I do notice on my Windows GPU Observer gadgets that once in a while the gadget readings for the second GPU will show zero for a bit and begin working again.

If Brother Matrix is having similar behavior with 6990M, I do wonder even more so if it is an AMD driver bug. It would not be the first, LOL.

 

I think this might be due to the new AMD technologies such as zero-core or enduro.

 

Whatever it is, I don't really care for it. I wish they would not put their tree-hugger garbage on enthusiast hardware. They should save that sort of nonsense for ordinary laptops and netbooks, and people that benchmark their batteries, LOL. It's not thrilling to see NVIDIA starting to do engage in such silliness, too.

Edit: zero-core and Enduro would not explain the same behavior on 6990M.

 

Not sure how much this applies since I don’t have an M18x, but perhaps these observations will be worthwhile:

I have been swapping MXMs a lot recently in my M17x R2. In order to make that easier, I haven’t been buttoning up the laptop. I didn’t screw the MXM in, didn’t put the black plate over the cooling fins, didn’t screw in the keyboard or pop top plastic control panel into position. Apparently some or all of that is necessary for thermal management / air flow. The 6970M I have in now was idling at 62*c that way. With everything installed and buttoned up, it’s idling 11*c cooler and the fan is MUCH quieter, maybe running 500 RPM slower.

Moral of the story – the hardware config is designed for thermal and airflow management. Put it all back and button it all down when you’re done.

Hope this helps someone.

I have to disagree with you there. My big laptops rarely have to be running flat out. I have never played a game on any laptop (heresy, I know, particularly since I have some really fast laptops). I have another setup for that which is a lot more capable. A lot of what I do with this laptop is code development, office apps and Blu-Rays on long flights. Since I don’t want to carry this beast AND another thing with a large display, I still want my Alienware to run for a 6 hour flight so I can code and be entertained. The new power management systems, particularly in the 7970M, make that (more) possible. Plus if I can save some power, as long as I can do what I need when I need to, I'm all for that. It just has to work right.

>90% of the time my CPU is clocked as low as it will go and so is the GPU. Not so when I’m benchmarking and giving demos.

Lest you might be lead to believe I’m one of those tree-huggers, My Evo VIII bumper says:
29 MPG Highway
19 MPG City
4 MPG Racetrack

 

I'm not even gonna start to pick holes in that........

 

Bring it!




Blah Blah Blah... <-- need to exceed 10 characters...

 

Firstly, I didnt mean for my comment to sound antagonistic, I just feel quite strongly on the subject of "power management technology" - just my humble opinion.

I appreciate you want to make the most of your machine whilst travelling on battery, so from that point of view I agree that power management plays a particularly important role, but if all you do is run your machine 90% of the time on uber-low clocks, why the heck did you get it? - I mean there are far cheaper machines that would allow you to do what you want with longer battery life to allow you to watch a blu ray on a 6 hour flight.

You may well have "another machine" at home (probably a high powered desktop, right?) that you do use for intensive activity but there are alot of M18x owners out there (including myself) that do not have another system for such tasks. I didn't spend the equivalent of $7500usd on a high performance machine to have it's legs cut off simply because AMD or Nvidia think they should run more "economically". If I wanted battery life and economy, I would have bought a more suitable system.

For people that use their machines in the reverse context (90% full power 10% energy saving mode) to yourself, they actually WANT them to perform at thier fullest potential, regardless of power consumption. These "power management" technologies by AMD and Nvidia just interfere with the ability to do that. I've lost track of the amount of times that AMD "PowerPlay" and Nvidia "PowerMizer" only end up causing more headaches for those who want "flat out" performance, and if even more of this "tree-hugger save the environment" crap is to appear on machines that are essentially desktop replacements, it's only gonna make things worse for those who do actually want to utilise their machine "flat out" the majority of the time as opposed to using on battery for exteneded battery life.

At the VERY least, both AMD and Nvidia should provide us with the ability to turn those power saving features off COMPLETELY so they do NOT affect performance.

Put in another context, these technologies would be tant amount to someone putting a speed limiter on your EVO - if you took it to the racetrack, you wouldnt be best pleased if it didnt perform flat out like it was designed to run.

I think what Mr. Fox was intimating was basically those people who want longer battery life and great power management technologies for run-of-the-mill tasks, they should get a system that is more suitable to those tasks and dont force these ridiculous technologies on machines that were really not made with efficiency as a primary concern.

 

> if all you do is run your machine 90% of the time on uber-low clocks, why the heck did you get it?

Benchmarking my code when away from home / office. I'm one of those crazy nutjobs that wishes AMD didn't depricate CAL. Gimmie back GDS please.

We have vastly different con-ops, both of which are valid. For instance, I won't buy an M17x r1 because the PCIE bus is crippled at 1/4 of what it should be. r2 is better but still crippled. This doesn't affect you nearly as much as it does me.

My ideal 'laptop' with current technology would be:
M17x chassis
17" retna display
Sandy Bridge 2687W
x16 Gen3 to each of 2 MXM sites
4 channels of DDR3 2800
3 or 4 SATA-III (or better yet, PCIE-attached) SSDs in RAID-0

But that would need a creative thermal solution and some serious power management. And would cost North of $5k. Like Aulavik North.

I'd still buy one.

> At the VERY least, both AMD and Nvidia should provide us with the ability to turn those power saving features off COMPLETELY so they do NOT affect performance.

Easy: edit the vBIOS p-states and set them all to max. Just don't complain when your GPU never dips below 65*c and fans are always on high and loud.

But the power management features should just work. 4-5 watts at idle, >>75 watts while gaming. If the GPU isn't running at max clocks when you're pushing it (gaming, or the stuff I write) then the power management is broken and needs to be fixed. On that I think we both agree. But that's the behavior I see when benchmarking. You don't when you're gaming?

 

I get where you are coming from totally. I was basically trying to say that having a third party application/program etc such as PowerMizer or PowerPlay on systems which are really aimed at high performance is kinda ridiculous. I get ya when you say about altering the p-states, something which I have done but think should not be required to ensure max performance. People dont buy Ferrari's or Bugatti's with such elaborate tree-hugging garbage in place so why should an elite power notebook be any different?

I still maintain that this kind of crap should be kept as far away as possible from enthusiast performance machines as the majority of people who buy them want exactly that - performance without restriction. All that PowerPlay and PowerMizer do is add to the problem of getting the very best from your machine.

But as you quite rightly say, we probably want different things from our machines - so with that, I totally respect your viewpoint and feelings on the subject.

 

Stevie,

All these "nannies" are on the machine because 90% of AW buyers never enter this forum. The majority will never remove any part of the chassis, wont even use compressed air to clean the vents (let alone a tear down) and they won't even know that playing on their "lap" is constricting the thermals. I believe AW does this to appease the majority who would be else be using their warranty way too much. Just my hunch on the situation.

Back on topic. Can anyone with 580M's with normal cooling let me know if 75*C under full gaming load in warm weather and 80*C during hot 90*F weather is OK for temps? They were low 70s under load in Winter. Don't know if it's just ambient temps or I need to do a tear down. Thanks

 

Dave, your temps seem pretty normal to me. Max temps I've seen on my cards under gaming load is about 72c with ambient temps around low 70c - that's living in the UK for ya!....your temps with hotter climate seem perfectly acceptable.

 

Thanks brother Stevie! Game on then.

 

Dave, I agree with Steve... temps seem about right to me, too.

I could not agree more. +1 Rep.

@BonsaiScott - it's fine to disagree with me. I still stand by my comments... you have a different purpose for your system than I have for mine. A single-GPU configuration is not something I am interested in. It represents compromise... something I don't often consider. I just don't give a rat about power management. I want pure horsepower and don't wish to compromise any of it in the name of battery life or "green" nonsense. None of the eco-friendly stuff matters to me even a smidgen. If I can accidentally get decent battery life and low power consumption, that's great... an unexpected extra is always nice. I run my system with CPU c-states disabled, full power on everything and every power-conserving setting I can find disabled. As long as temps remain in the safe zone, I don't allow my system to slow down for any reason 99% of the time. I also travel extensively for business with the M18x at my side. I generally find switching to the IGP is enough battery life for me. It's nice to be able to extract 4+ hours of use from it, but I'd give up that ability in a heartbeat for more horsepower.

 

Guys, my system havin strange fan issue with GPU 1, as after fresh install Gpu 1 fan starts spinnin when temp hit to just 50C, and its keep spinnin even at temp 29C. It never happen before, I have tried different driver but not any luck.

 

One of my 580M graphics card at idle is 60 ° C and the other at 45 °. Is it not normal right?

Yesterday I changed the thermal paste of the core, and thermal pads of the memories just broke a little bit (just a couple, and not much) I'm sure this is what is causing the problem.
What thermal pads should I buy and which is the correct thickness?
Wich are the best thermal pads on the market for this notebook?
Do you think it's better to call Dell's service tech and ask for a replacement?

Thank you

 

It's not odd to see one card higher than the other like you say. Is this whilst browsing or just when sat at the desktop completely idle? - if its the former, that is quite normal to see some "cycling" activity where one heats up as the other cools down.

 

Do you guys run your fans on max through HWinfo? I would imagine running fans full blast is better than letting the GPU's heat up? Fans should outlast a GPU right? I just manipulated the fans to blast much higher and my temps are much lower consequently. Is this safe and good to do?

 

Well there is no need to kick fan at higher RPM, additionally you will get annoyed at long term of use. Fan usually runs higher when it needed, heatin GPU around 80C its fine, even i dont think 580m easily heat up to 80C.

 

My 580M SLI setup never got warm enough to require the fans to be revved up. I only use HWiNFO64 to run my fans at 100% for benching and that has always been for the CPU being overclocked to like 4.7 to 4.9GHz. Otherwise, there was never a need.

 

Same here. I dont use manual fans. I havent done much benching of late so I leave the fans on auto. I am yet to see my gpu's get anywhere near 80c under normal gaming use.

 

Fans on auto here as well. Highest temps I have seen are 78C in a hot room while overclocking. Mostly I see 72C max, 63C max with VSYNC on for most games.

 

Today i have noticed some wired behavior Gpu 1 fan profile. As it startin spinnin at noticeable speed whenever temp. hit 50C, till it reach to 35C. After reachin 35C still it wont rest for next approx 5 -7min.

i was workin on M18x stock display, i had use it almost more then 1 hour but all of fan was so quite,ive check Gpu temp., it was about 62C max but even tho fans were silent. Later ive connect system with my U2410 monitor, again fan issue starts popin head on top of horse, as it kicked on when Gpu temp reaches 50C. Hence ive figure out that fan profile is bein destabilize when external monitor connected to system, either HDMI or Mini DP.

 

Peter, my fans are doing stupid things also and I don't use an external display. I'm not sure how that would affect the system. If the fans were to only operate properly with an external display it would be a very crippling problem. I know it is not an option for M18x R2 owners, but I dropped back to the A03 BIOS for the M18x R1 just to see what, if anything, might improve.

I posted this earlier today, before flashing A03...

 

Mr. Actually my problem is with external monitor, with internal display all fans stays very silent at normal use, its never kicked on, even tho if im watchin youtube videos that i have noticed that core clock rise to 450Mhz and temp reaches at 64C then also fan stays silent. But when ive connect external display, fan start spinnin as GPU0 temp. hit 52C and stays on even at 32C temp for approx 2-5min more, same cycle repeat at temp reaches 52C. I have no idea how come fan profile destabilize when external display connected.

Yes you have R1 so your able to check system with different bios but with R2 there is only one bios release so far. As i remember you will face thermal shutdown with A03 cosz of unstable fan profile.

additionally GPU1 will stays warmer cosz it fan wont spin till temp reaches to 65C while GPU0 fan usually spinin all time a very low speed.

 

Apparently there is a minor difference with the external monitor. It may demand a little more from your GPUs than the internal display. If you are driving 3 displays that would likely be more demanding also, especially in terms of video memory utilization.

I haven't seen any issues running the M18x R1 on A03. So far, no change relating to the video cards has been observed than I can pinpoint.

 

Yeah, there is a slight increase in temps when using an external. Probably moreso if you use the internal AND external at the same time. As Mr. Fox said, the more screens getting run puts a greater demand on the gpu's to "drive" them. Here is a HWiNFO shot of my gpu temps at idle (well, chrome open, bit of browsing) when running only the external:

 

Yeahh agree with you as external monitor demands more from GPU, but i dont think its issue in this case. As today i have use M18x on stock display long time but fan wasnt kick on, even tho i was watchin youtube videos so core clock were runnin at 450Mhz, highest temp noted was 65C but still at that temp fan stayed silent.

Now i have connect my monitor, started watchin youtube videos so temp started increases from 37C at ideal, when it reaches 52C fan started spinnin So ive really dont understand logic, with stock display it wasnt kick on till 65C(as max ive reached while watchin video at 450Mhz), but it does with external monitor as long as it hit 52C mark, additionally it continuously runnin even at 33C for more approx 5min.

But fan profile is related to temp. isnt it? I have noticed that with stock display or external monitor temp. increases in same manner. Fan stays silent up to 65C with stock display while it kicks on at 52C with external monitor

Also i have noticed that sometimes neither Afterburner nor HWinfo able to monitor GPU 2 temp. :S

Below is my system current temp.

 

I just repasted one of my GPUs (my primary - secondary is inoperable while I wait for my new SLI cable to arrive). Specs in signature. My temps seem to range from 55-to-low 60s on idle and mid-70s (never got over 76) under load (tested while playing SWTOR). I'm using an external display as well, and I do not have a cooling pad (it's on order). Are those temps okay?

 

They seem pretty normal to me, Mike - what's your ambient room temp like? - you may be running a tad warmer at idle than me but I'm in a rather cold climate, so if your warmer, those idle temps seem ok. Load temps seem perfect....are your idle temps without anything open at all and jusst sat idling at the desktop?

 

for anyone interested in 1 pipe vs 2 pipe heatsink difference on a 3720QM you can read about my experience here : http://forum.notebookreview.com/ali...-cpu-overclocking-thread-128.html#post8719041

 

Finally received heatsink , AW tech was too scared to fit lol

 

Too scared?? LOL - that's kinda funny in a not so funny way!!

 

i just put my baby on the desk wiped off the speck of dust from lid " and said you have worked on one of these before havent you " Face said it all lol

 

lol, I prefer to do this kind of stuff myself. It was not that hard, just have to take your time. I did not expect that I'd have to remove the screen though.

Will be interesting to see what kind of temps you get with a 3820 and that 3 pipe.

 

I will post some screenies later , i am hoping to get some good temps with this heatsink

lol yeah i wasnt too impressed removing the screen of my new m18x ! was a bit dodgy had to change the pants a couple of times lol

highest i have seen so far is 55 degrees , benchmark SF4 and SF vs Tekken and 3D mark 11 . i would highly recommend the 3 pipe HS !! man i love this laptop

 

55 degrees after 3Dmark11?

 

what settings are you using to overclock the 3820QM ? i wanna give it a try ?

 

Through bios: 3720QM or 3820QM overclock | Понятные вещи - для понятливых людей!

here's the translation:


1. Flex support disabled
2. Extreme edition Enabled
3. Level 1-3 overclocking - level 1
4. Override turbo settings enabled
5. Increasing core ratio limits to 41, 40, 39, 39 (if 3720QM, make it less by 1 in all 4 cases)
6. Flex VID override - 0
7. F10 - yes

PS Post your temps after successfull run of 3Dmark11 and prime95. Thanks

 

ok will give it a try dude and post back

 

Where's the report?

 

lol !! got side tracked last night i will do one today for you

 

Maybe if you're not using so much ur new 3 piped heatsink - u'll send it to me?

 

lol have you not tried ringing AW support and sourcing one off them ? i got it for free i just told them my temps was high and i told them the heatsink i wanted and bam ! on my doorstep yesterday

 

I call them every week. They are telling that they are out of stock.... And i am from Kazakhstan, not US I just ordered a 2 piped heatsink i think there won't be so much difference between 2p vs 3p. When you can post your temps?

 

when i get home tonight mate , as i am at work ..