Safe Temperatures for nVidia GeForce 8600M GT

right now I am watching my nTune monitor and it is saying that my GPU is running a 88 degrees celsius. It was going up to 92 earlier. Is this safe? The program is defaulted to display a warning at 127 degrees, but judging by what I have read in the forums it should be around 60 right? What is going on? It's not even overclocked, and all I am doing is talking on AIM...

 

try not to use the notebook on a surface that blocks the fans. many ppl report higher temps than yours and say they have no problems but u might want to buy a ccooling pad

 

try another program like rivatuner and see if you get the same temps. 88C is way too high though doing things like that, is anything blocking the vents? Does your laptop feel hot to the touch?

 

according to rivatuner, my 8600m GT runs at 41*C idle, and thats without a cooling pad (although i ordered one off ebay yesterday )

according to i8kfangui (google it -.-), and if i force my fans to high using the same program (not recommended if in game, gives you this really annoying lag [see other threads started by me]) it can get to as low as 35*C idle, without cooling pad.

so i would say that 88*C is a pretty alarmingly high temperature for idling.... but it depends on how big your laptop is and if the cooling is sufficient as well........


edit: ntune monitor reports the same temp of 41*C idle (although its getting lower now - forced fans on)

 

Under Rivatuner I Idle at 47, with a +7 offset (rivatuner is off for me)

 

what could be the problem? my C90 has 4 huge fans and and it can't be cooled? I have the bottom off with the computer elevated, the fans were on during game play which still pushed it up to 98 degrees, and I now have a 2x2 foot fan behind it leveling it off at 57.

 

surface mount transistors denature and melt at 110C and up. I had a transistor blow on my Quadro, when the temps were only 38C. It happened because the card was defective. Prop up the back of your laptop so that air can circulate, you will notice a drop in gpu temps. The cooler the better, I know once my gpu hits 60C, I have riva tuner come up and warn me that my temperature has reached my personal thermal limit. SO after an hour of Crysis, my gpu will hit 60C, and at that point I shut the game and get back to work.

K-TRON

 

My friend Kate has a C90S, and it worked great until she came to me because she burned herself in many areas. Literally I tried with that thing, and the temperatures would not decrease. Kate plays wow all of the time, and her gpu would sit around 100-115C. So after 2 months of owning it, she had to send her laptop back to asus for a new graphics card. Yeah, things burn out but that is only because ASUS laptops have cooling issues.
K-TRON

 

This is a message I sent to Kate, before her nvidia card burned out. We emailed tech support, and here is what they said:

This is my message I sent to ASUS:

My friend Kate owns this laptop, the C90S, and she is not at all tech savvy. She told me how hot her laptop was getting. To monitor the temperatures of the processor, we installed core temp. To monitor the graphics temps we used riva tuner.

Core temp states that the processor, the E6700 @2.66Ghz runs between 55-70C under normal use. However, by after just a few minutes, the GPU, nVidia 8600GT 512mb card reaches 70C. Under normal use, her gpu runs between 80-95C. When she opens a game, her GPU temperatures shoot
up to 110-120C. The games she plays are portal and world of warcraft.

After noticing how hot it gets, we upgraded the Bios from version 0601 to 0801. Looking through the bios, under hardware monitoring, you can see the cpu temps and fan speed. The four fans in the system, are 12volt fans which are designed to run at 9500 max rpm's. The bios states that they are only running between 1800 and 3700 rpms.

My question is:
Is there a program which can force the fans to full speed on your ASUS C90S?

If these temperatures continue, her laptop will burn out pretty quick.

The turbo gear program only lets her laptop run at different clock speeds. There is no fan control in the program.

There is no dust in the system, and even if we prop up the back of her laptop for more circulation, these high temperatures continue.

The answer is not because its a desktop replacement, because I own a more powerful Voodoo desktop replacement, and my computer never hits half of those temperatures.

Can you please tell me or give me a copy of some kind of fan control program, as she needs her laptop and does not want it to melt.

Any response would be greatly appreciated

Christopher Kouttron


Here is what ASUS tech said:


Hello,

Unfortunately, no utility software is available to control the CPU fan
speed. Since the BIOS was upddated, try loading User Defaults to
re-aligned the CMOS mapping for the newly installed BIOS. Verify if
the heatsink on the CPU and GPU is installed properly and apply an
ample amount of thermal compound to inprove heat dissipdation.

Regards,

RL
ASUS Notebook Technical Support


This is what I emailed Kate after the response:

SInce ASUS doesnt even know how to fix their own laptop, I had another
idea. What we can do, is find the thermal probe. WHat we can do, is
solder a copper line from your GPU heatsink, and have it touch the
thermal diode. Since your gpu is what is getting hot, the heat will
transfer through the copper line, reach the diode, and cause the fans
to spool up.

If you dont mind, we may have to open her up tomorrow afternoon, to
see what we can do.

Using that huge fan is ridicuolous, its better if we fixed it.

K-TRON

She still has not gotten her laptop back from ASUS, so the adventure continues.

K-TRON

 

@Shpongled:
Using RivaTuner I have 50°C at idle (just web browsing and chatting). My system is a Vostro 1500, no overclocking.

I would suggest to be careful since those values at too high even among people that have overclocked their cards. Do you have your card overclocked??

Edit: And when running 3dMark06 the temperature increase up to 61°.

 

No I do not.

 

The C90S has a lot of thermal issues.
Shpongled, since you have a C90S, have you took the bottom cover off?

There are two cables which connect the fans to the motherboard. What I think is causing the overheating, is that ASUS accidentally plugged the cpu fans into the gpu fan connector in the motherboard, and vice versa. It is worth a shot to try and switch them, because some C90S owners have overheating issues, and others do not. All of the fans are thermally regulated by an on board sensor.
If the wires which power the cpu an gpu are switched, then when the therml diode is activated it should turn on the gpu fans, instead of the two cpu fans.

I dont have Kates machine in front of me, but when I do, which will be most likely in late january, I will be able to figure something out.

I told Kate, that if I cant fix the overheating issues, I will have to make a PCB, which takes the 16 V from the screens backlight, and convert it to 12 v, so that the fans will always run at full speed.

K-TRON

 

interesting idea. I don't know which is which though.

 

Its simple, there are two groupings of wires, one group from the two fans which blow on your gpu, the ones on the left, if the front of the notebook is facing you, upside down. The other grouping is the two right fans which are for your processor. Bothe sets of cables connect to the mainboard, never the bend in the cpu's copper heatpipes. You should be able to swap them, hopefully that is the fix.

K-TRON

 

Hey guys,

Another important point is to consider is whether you are using the correct GPU drivers. Although most don’t pay much attention and install drivers form third party websites, this could be really harmful if the driver bus/memory speeds are configures to use a higher settings rather than what’s really available on the hardware…

If you have upgraded GPU drivers from the original laptop manufacture’s drivers then rollback to the originals or the latest from the laptop manufacturer and see whether it makes a difference…

Good luck…

 

yea I reverted back to to the original MXM drivers, which work MUCH better than any of these drivers on laptopvideo2go.com, but the temperature is still the same. I will try your suggestion when I have motivation. Im taking a break from this right now.

One good thing I have found though is that I can now safely say that one of my previous problems (screen flicker and colored specks) WAS in fact due to an overheating GPU, because when I took off the bottom of the computer and put the 2x2 foot fan behind the laptop to cool it down, those problems stopped.

could this maybe also be due to something wrong with the thermal paste, or does that not have a significant effect on temperature?

10 minutes ago I unscrewed a metal plate that had a copper wire running to what I think is a heatsink, but definitely related to the fan, and tightened it. I think the plate and such was connected to the GPU, but I didn't want to mingle around anymore for fear of breaking something and voiding warranty. But now my computer seems to be leveling off and stable at 76 degrees Celsius, a little hot, but 30 degrees cooler!

Are the fans supposed to shoot hot air out, or suck cool air in?

 

If the laptop is under manufacture warranty, get them to solve the problem for you. If you are sure of that the drivers are not the reason for this overheats then it has to be some sort of a manufacture defect in the cooling system.

 

I'm not really a fan of sending my computer away for 2 weeks while I am on winter break from college.

 

Actually your statement is totally wrong. Installing/updating drivers does not change the core/memory frequencies at all. What they do is how to manage the information, that's all.

 

Seconded. It's the video BIOS that controls the clocks not the drivers. I once instslled an geforce 6600gt driver on my old 6200 and it not only worked, but the clocks remained the same. Otherwise it would've probably fried the poor 6200

 

are those temps at idle or at gaming? if they are at idle u have a prob, if they are at gaming then it's ok. i know it seems high but i have the same GPU in my acer and when gaming it runs at the same temps. at first i thought it was kinda high too but after reading some internet stuff i found out it's with most of the 8600m GT gpus. I even have it OC'd now and it stays at the same temps.

 

for lazanogo -
i'm sorry i really did not mean to say that the drivers control the clock speeds etc, but i do believe the information that contains in the drivers directly impact the clock speed etc... correct me if i'm wrong. i guess thats why some drivers (3rd party) which claim to work with a particular GPU model does not work well as one would expected.

for bubba_000 -
nVidia drivers are universal thats why they work with every GPU. drivers are aware of the GPU installed in your system thus they use the correct settings during the installation.

i did not know that BIOS hold the info about the clock speeds etc, thanks for that information.

cheers !!!

 

well like I said, it gets that temperature even when just browsing or talking on AIM. So its basically idle.

 

I'll just get my two cents in there:
Safe temperature for the GPU does NOT, I repeat does NOT, mean that the temperature is safe for the LAPTOP!

While it's probably true that the GPU's silicon itself can withstand 120C or so, can the OTHER parts that's near the GPU withstand temperatures of ~100C? I don't really think so...

 

The information of the drivers does not impact the clock speed. Their purpose (for releasing new drivers) is to improve the information management (in general or specific games) or to avoid possible known crashes/issues with the specific card. In normal use, the clock speed will vary depending on how much work it has to do, i.e. the clock speed will be low if you just web browse, but will work at max while gaming if the game really needs to pull out most of the resources from the card.
The drivers are universal in the sense that they apply to all the line of cards a specific company, i.e. nVIDIA, but if you read the release contents you will see the changes they do to each specific card.

 

I just played hl2 episode one (i bought the orange box ) for over an hour and a half and my max. temperature i had was 93°C (WITHOUT OVERCLOCK: just the standard clock settings)

 

I addition to what I reported (showed here as quoted), after running Oblivion for more than 1 hour, the GPU temp can peak up to 69°C, averaging at high use around 66°C-67°C.

 

My Zepto 6224 with Rivatuner reports around 97° when gaming. After a standby, which somehow effects when the GPU is down throttled, it reaches 111°

 

okay, it is not that i disagree with you but how do you explain driver tweaking? websites like laptop2go.com says on their own website that they modify the INF so that hidden information such as over-clocking can be done ???

 

Give a call to Asus, perhaps they'll be able to give you a replacement or solve your problem. IMO, that's pretty alarming considering that's almost zero work for your GPU to get to that temp.

 

I am not sure what you are talking about. If you can provide a link would be very useful. As far as I have read the modified .inf file is just to override the driver installation on the laptop, that's all.

Also, there are tools that allow you to overclock, but that has nothing to do with the driver installation. Such tools are nTune and RivaTuner (maybe other, but I don't know them).

 

Well a driver drives the hardware. It tells the hardware what to do. The driver allows a GPU to be or not to be overclocked, it unlocks the clockspeed. Then u have tools like nTune and RivaTuner. These are indeed used to overclock the GPU but basicly they tell the driver, to tell the GPU to run on a higher clockspeed. So it could be possible when he uses a homebrew driver that someone managed to made wich already change the clockspeeds. Only that is probably not the issue because as far as i know it these drivers are nog distributed.

But so the TEMP you are getting is @ idle, in that case u need to have to call support. Maybe your cooler is broken, anyway there is something wrong.

 

I am totally looking at that right now. Dude, if you're right, you're my new God.


EDIT: It looks like they were connected like they should, black on black, white on white, when I put pression on the cables you can see it on the good FANS. Since my CPU is running pretty cool, I still swtiched them and I'll look at what it does.

EDIT2: No differences, except that now I can hear the fans while gaming (and god I like that sound) There is 3 CPU fans and 1 GPU fans. It seems like the 3 CPU fans are missing some power. From what I'm hearing right now, I say the problem lies with the GPU fan, not the sensor

 

All right I have a FL90 with a 512 GeForce 8600GT. Now it seems to run fine, a little warm though.

I installed Oblivion. When I play this for about an hour or so my Laptop seems to get pretty warm in some places. So I intalled RivaTuner (i think that is what it's called). On Idle my GPU runs at about 60 C. I turned on the in game view now after about an hour of game play it will reach a temp of about 80-84 C. I was concerned about this so I called the tech support for my laptop. They said that the GPU temp runs from 70 to 90 during game play. Is this normal??? I feel that may be incorrect or can GPUs run at these temps and still perform correctly.

Please reply!