Windows 7 GPU Overheating, please help

Hey everybody, I have the G50vt-x1 laptop

UPDATE:

Hello everybody, for those of you who just started having this problem, let me quickly say that DO NOT Play with the insides of your laptop, as they are not at fault.

This seems clear to me now that it is a problem by the final versions of Win 7 (Currently running Professional). I had RC and never had any problems.

Windows 7 might be "overusing" the hardware, so the system seems snappier, a cheap trick, but I think that could be it.
I personally found little to no improvement on speed from RC to Final, so "over usage" might not be the problem.

As you can tell, I have no clue either...

In the beginning, the game would run for a few mins then become extremly choppy, this was due to the GPU hitting 110C, which then makes it underclock greatly to decrease in temprature.

My second problem is the CPUs hitting 100C as well,

Everyone pretty much said that my gpu & cpu running at 100+C is a death wish

See pics here http://forum.notebookreview.com/showthread.php?t=431512&page=2


I suggest to stop with the gaming/3d applications if possible because overheating and cooling down will certainly wear out
the chips.

I have been doing all kinds of test with still absolutely no clear answer on how I was able to make my gpu to only hit up to a 100C, which enables smooth gameplay.

So far, all I can say that this problem is only isolated to 3D applications, as my laptop, if I cold started it and ran it with no 3D apps, the laptop would run at very low temps, but if a 3D game was recently played, my gpu and cpu idle around 90C/80C

Original Post

After the installation of the full version of Windows 7, I started having overheating issues with my video card.

I ran some 3D tests and when it hits 110C, it downclocks so much that things become unbearable.

On NON 3D applications, it hits 110C+ and still no slowdowns what so ever, I've had it hit 130C when I had an 1080p HDMI out to a TV playing the simpsons.

I had the newest official driver, then the BETA driver, and now on the Asus website driver.

I'm running out of options, I didn't have such a big problem with overheating on Windows 7 RC, and it was running fantastic on Windows Vista.

Please no recommendations for coolers, or stands, i've been using this laptop over a year and never had such a problem, and I've already cleaned my laptop.

I need real help with this.

 

Damn. 10char

 

Are your fans spinning?

 

yeah I've got 1 fan in the laptop, it spins the same speed it always has been, 6000rpm

 

And you haven't touched your internals?

 

I did an upgrade right when I got it, I upgraded it to the T9600...

But this didn't affect my temperatures at all in vista, I just had a great speed increase.

Currently the 9800GS is at 89-88C
and my processors at 60C

with nothing running but firefox, msn and Everest (the temp reader)

 

download hwmonitor, I've had some rather glitchy temps from everest.

 

And you are sure your fans are running? The only way I can imagine getting 110C on your GPU is if your fan is borked.

And I think his overheating is real, since he's downclocking.

 

I really couldn't have cared much if it was running perfectly, but games and 3d apps are usable.

My fan is definitely pushing some air out, I opened the laptop yesterday, I cleaned it (there wasn't much dust).

There is 1 heatsink leading from the 2 cpu cores to the fan, then another heatsink from the gpu to the fan, and that's about it.

I feel this is not a mechanical problem but more a software problem.

 

at 114C his GPU is dead (according to nvidia) so 130C sounds unlikely to me. His GPU would've been long dead by then.

 

Trust me, we know what the innards of a G50 look like.

As for your fan, does it ever spin up or spin down? Is it going full blast during boot up?

 

I'm not making up numbers, I got 3 apps to prove it to ya.

I have everest, gpu cap and now hwmonitor...

Will post pics soon... at 90C... it climbs up pretty quickly

 

Yeah it's going full blast on boot up, then goes to 4500 rpm... then with slight use goes to 6000rpm...

I don't recall getting to 6000rpm with minimum use of the laptop with vista or even 7RC.

 

Have you replaced your paste recently?

Software shouldn't be putting out those temperatures. Even utilizing 100% on both CPU and GPU you shouldn't be getting those temperatures that quickly.

 

I never have replaced the paste.

 

Then I honestly don't know. The only thing that should be wrecking your temperatures this badly is a bad fan and/or bad heatink interface with your dies.

 

i would probably roll back to Vista/Xp, just idling at 90C its really dangerous not to say the 130C, that wierd it should have a thremal downcloak shudown, but idk maybe win7 is producing the buggy temps. Try to reintall if you can or roll back to winxp/vista and see if the temps are the same.

Btw how are you reading the fan rpm, i havent found anything that will read or control or even monitor the fan rpms.

 

everest, reads fan speeds... I tried finding something to control them... even in bios... it was hopeless

 

the bad thing about this, is that I don't have a vista cd, i'm moved and I lost it.

 

Here is a pic... it says 113 on all apps.

It's currently on 115C

 

By chance have you opened up your computer to check that the heatsink is still touching the GPU?
I've overvolted and overclocked my G50vt-X1, and I run GRID for hours at a time and I have to really be pushing the GPU with moderate ambient temps to break 95C. I feel like if you're hitting anything that high with the fan on high there's a physical problem with the way your heatsinks are lined up, or something is not transmitting heat correctly. Ensure there's no gaps along your heatsinks and consider re-applying paste.

 

yesterday I opened it up, I tried to remove the heat sink, it was too well connected, I just tightened the screws and closed it back up.

 

That'll do it. There's your problem.

 

But this problem didn't spring up yesterday...

But tell me what can I do then.

It's now 118C

Update...

Shut it down for 10 mins, checked the connections to the GPU and CPU.. very well, I'll need to add thermal paste to the CPU cores... but I can't even remove the heat sink from the GPU... it's too well glued.

Turned it on, my comp started with cores at 60 and gpu at 65

 

I replaced the GPU and the CPU paste, I did a good job it seems because the GPU is now taking a slower time to climb up the temperatures, but it's still idling at a ridiculous 89C, and still reaching the sky high temperatures....

 

Well try to see if the cooler is fully going down, when i change paste the first time i though i assembly it back alright, but the radiator is tricky going in made it not seat correctly into the gpu die, i was getting 70s idle, i re did it again, wasting some paste trying to do samples into how much was good enough to make contact, and it work alright at the end.

 

Well I got it fixed, I wish I can say how... I installed the 186.03 driver (an older one, thanks for the tip dtd00d), uninstalled a few things, installed direct X, vacuumed the laptop, re-applied the paste.

Now when playing Forged alliance, my GPU hits 90... and stays there!
my 1st core hits 100C and my 2nd core hits about 90c

Very happy, extremely stable... within seconds of quitting the game the temperatures go down drastically.

Thanks for everybody's help


btw Abula, I know what ur talking about, first time I did it, I spread the thermal all over the place when I was trying to place the fan and heatsink back.

 

90C for the GPU is...okay, I guess. It's a 9800m GS, so it shouldn't be doing the wonky temperatures of a G51, but oh well.

But 100C on the CPU? That's just bad. CPUs epik-phail at, what, 105C?

 

While 90C still high i think you did install it corretly, at least not going for the 130C that was before. On the cpu is not good though, most of the monitor programs have 100c as the upper limit, as forge said, probably 105C will start doing permanent damage, so you should try to fix this, since the gpu temps seems more normal, my guess is that either the cpu part of the heatsink isnt fully seated or too much / to little thermal compund, usually dual core shouldnt idle higher than 50C and should be lower than 80C on full loads (this is with stress test, in game maybe 60-70s at the most), the heatsink is semi independant (one part for cpu/nb n one for the gpu), so try just to see if the cpu part is correctly seated and and touching both nb n cpu, maybe one of them is creating a gap to the other.

 

omgz i would be crapping bricks with that machine, i recently ran intel burn test on my Q9000 and i nearly cried...using 2890MB ram, it reached 83C all cores and it seemed like it was rising and i just quit, cuz im too dam afraid of the reprocussion of having cores at that high.

anyone know if there is ANY application besides Intel Burn Burn that can raise temps that high? i needy stay clear of thoes apps lol

and to ur problem, i think u should put that thing into asus hands, you have a really screwd up machine rite now. OR maybe try thermal paste on the CPU cuz it seemed to fix ur GPU problems.

 

Well, I played Operation Flashpoint 2, and my gpu reached 100C...
core 1 did 95~ and core 2 did 92~

Right after quitting, the tempratures go down quite quickly to gpu being 65~ idle and CPUs at 60~ idle...

The cpu temps drop much faster than the GPU, i thought this was worth noting.

I guess it could be the thermal compound, maybe i put too much or too little. I'll be putting thermal pads soon to see if a difference is made...

Btw, my fan is now going up to only 5500rpm... I think it used to go to 6000rmp, can anyone confirm?

This could be the problem now for my high temps.

 

Yeah sounds like things are going back to normal. There's no way you could have such high temps before other than if your heatsink wasn't contacting your gpu. Those temps are still a tad bit higher than normal I think, but definitely not alarming anymore.

Before I undervolted my GPU, it would idle between 53-58 if I remember right.

 

I did some more re-installations of drivers, I got the CPUs to be at 50c idle.
Gpu is still around 62~

 

This thread might be related in someway

 

Thanks, that seems to be connected... I'll check the to see if that is true... I thought aero was solely by the cpu...

 

hmm 50C seems good but no way in hell should it rise 42C just by a game, specially flashpoint. if it doesnt go to 92C, say it reaches @ max of high 70s, i think you have an operable G50, altho not in very good condition.

 

Oh man by those standards I think I'm in trouble, too. I idle at 45-48C and hit 95C whenever I play GRID. I hit 97 once, but no higher. That's a 50+C increase but my G50 is holding up great (so far).

 

NO you are fine, that is how the G50 was designed. it isnt abnormal, most pre-modded will be like that. i believe G50 get slightly lower temps but due to u owning it for a longer period, temps could be caused by dust etc. g51 has a higher watt TDP gpu, i believe its 60-75 or something, and is much hotter than the 9800m gs in the G50s. i have nearly same temps as u, maybe +1-2, havent played grid in so long. that is with overclock tho, on stock my temps stabilize @ 90C.

i was talking about 42C increase in CPU!!! that is a ton, NOT NORMAL. gpu is normal to increase that much...in asus notebooks that is.

 

in Superior Commander: FA, core 1 hits 100, core 2 hits 90~...

Still trying to solve problem... maybe some intel drivers now...

 

If your CPU and GPU haven't curled up and died already, it must mean you have faulty temperature reporting sensors or programs.

 

o.o good idea, maybe win7 + hwmonitor = fail? cuz srsly 100C on cpu...thats insane

 

HWMonitor works fine on Win7, I and many others have used it just fine. If anything is wrong it is on a hardware level- a faulty sensor or something worse.

 

Then it has to be hardware level. No CPU can survive those temperatures.

 

how does it explain the problem arising after the win7? two cases have been posted with basically same problems...overheating AFTER win7.

and yes i do agree, sorry for the foolish comment, a lot of ppl use win7 and they have hwmonitor..derhhh

but yea, new paste, reseat heatsinK?

 

Well finally my laptop started shutting down after hitting around 100C in my cores.... now I can't seem to control my CPU Heat, where as before I couldn't control my GPU.

This may not involve Win7 but rather the drivers from ASUS... because the new win7 64 bit drivers do not seem to be for my laptop specifically.

As I was typing I realized that the problem may actually be from the ASUS drivers as I upgraded my CPU, so the drivers are actually for the p8400 not my T9600.

Anyone else upgraded?

 

CPUs should be plug and play, no need for drivers.

To be honest, I don't know what the heck's wrong with your computer.

First we thought it was user hardware error, in that you might have done your paste wrong. But you said you replaced the CPU paste and still got hte same temperature readings.

Then we thought your fan might not be spinning, but you say it is.

We thought it could be faulty reporting on the software level, but Element has said HWMonitor works perfectly fine in Win7.

Then we thought it could be a hardware temperature readout fault, but if it's shutting down that's a sign of overheating.

So I have idea at this point. Granted, I don't really know a lot, but there you go.

 

CPU's are not plug and play. He still needs a screwdriver to lock and unlock the CPU, dum dum.

 

Okay, okay, you need screw-drivers to install it.

 

This is user error.

If the CPU HSF & GPU HSF is properly installed with thermal compound, there is nothing out that there that would push your CPU higher than maybe 60C degrees at 100% Full Load, unless you are running a CPU stress test.

Now that the GPU looks like it has normalized. A clear indication that you should not be working inside a computer if you don't know what you're doing.

 

No such thing as CPU drivers. Chipset drivers control how your CPU and Motherboard features interact with the rest of your system. Take it to someone knowledgeable and have them fix it.

Or not, since the way it's been running, I would not be surprised if the GPU and CPUs have been damaged.