The gtx260m in my g51 used to run at about 87c max under load for quite a while when i first got it. After the first month or so, the max temperatures started to increase a bit. Almost every week to two weeks now, the max temperature seems to go up by 1c. So, after a few months it is now at 97c max. That is the highest I've seen it reach (except for league of legends which reached 100c in about 10 minutes... cel-shading hurts? :X).
Now, you and I both know this isn't a healthy temperature, especially since it just keeps rising. I prop my laptop up with a book in the back, I just opened it up to check for dust, didn't notice anything, but still sprayed compressed air into it with no luck.
What do I do? Do you think it might have to do with the thermal paste or heatsink? (I'm not very computer savvy by any means, so when I opened it up I didn't check to see if something was wrong with those since I wouldn't even know if it was messed up in the first place). And if there was something wrong with something, why would the increase be so gradual? You'd think it would be a bit more sudden or something.
Please, any help would be greatly appreciated.. I still have that 1-year warranty from best buy, so no.. I'm not going to do anything to break it.
-
-
SoundOf1HandClapping Was once a Forge
1. Yell at it.
2. Try different video drivers to see if those are the problem.
3. Scream at it.
4. Dismiss the Worst Buy warranty and repaste.
5. Threaten it.
6. Make sure there's nothing wonky with your fan.
7. ??? -
Could be a bad thermal paste job and/or heatsink that is not making full contact with your GPU. If you want to try, you can replace the factory thermal paste with something better like Arctic MX-3. Do a test placement of the paste to see if full contact is being made. If not, use a thicker layer of paste until it is and then seal it. You should see improvements right away.
Also, you may consider wanting to check out the copper and drill thread started by forge. It talks about adding copper blocks to the heat pipes to further lower temperatures. Lastly there is the panel hole mod that can be done to reduce overheating even further. If you did a search, you would have found a solution right away. -
@joker no see, that's the thing. i know about these type of procedures, but i don't trust in myself enough to risk breaking the warranty. and i don't have a problem with overheating, i have a problem with my heat gradually increasing. even if i were able to bring the temps down a bit, whose to say it won't just climb back up to where it is, following the pattern? and i'm not sure i know what you mean by doing a test placement. is this something that would break the warranty as well?
@forge, i think my fan is acting the same way it has been. is there a real way to make sure it IS functioning correctly? a strange thing i noticed is that in the program that records various hardware information (similar to HWmonitor but i think it's asus-specific and came with the notebook (i uninstalled it not too long ago)), the few times i checked it, it had said my cpu fan was at 100% (both of these were when the notebook was idle, and it definitely was not as loud as it gets when gaming). just stating.
it might've started getting hotter once i got the 186.82 nvidia drivers... maybe i'll try something else? the dox drivers perhaps? even though i hear those run hotter :X -
My idle temps increased about 5-10 degrees after I installed new nVidia drivers. The gpu wouldn't go below 62C while idle. After installing the DOX drivers things went back to normal and the drivers work real well. GPU can idle at 51C and below now. Maybe give the DOX a try.
-
the new nvidia drivers do the same, i idle at about 48-52c. idle temps are great, it's just under load something must be going wrong.
i'm starting to think it's my fan. why? i don't think it ever goes on full blast anymore unless it reaches insane temperatures. also, this is probably worth mentioning. for a week or so, when i'd boot up the laptop, sometimes the fan would go on full blast and stay on until i shut down. that stopped happening. kind of weird, huh? well, the sound level of that fan was definitely louder than what I hear when i'm playing games that fully stress the gpu now. so.. i think my fan is not functioning properly. -
If your fan kicks on full blast at start up.... i had that happen to me. i used nvidia system tools to raise the clocks and my laptops fan would kick on from the time to i turn it on to shut down. maybe somehow the clocks got changed?
-
Or maybe the heat sensor? Perhaps give Asus tech support a ring. They might have dealt with this before. I generally keep my Vx on a cooler. I still don't like to see the temps go up while watching movies or surfing with music on. For gaming the cooler is full blast.
-
SoundOf1HandClapping Was once a Forge
It's 66.9F and my GPU while surfing and watching movies is 44C (Min of 40 and max of 48).
Anyway, the main reason for high idling temperatures is that the GPU is not downclocking via powermizer. Try to install powermizer separately or try different drivers.
And, like I said, the Worst Buy warranty is a laughable warranty. Just paste it yourself. One out of one Jokers approve -
CitizenPanda Notebook Deity NBR Reviewer
Just use some compressed air and clean out the fins and fans first - then come back and report. You can do more drastic measures later.
-
-
CitizenPanda Notebook Deity NBR Reviewer
-
i sprayed pretty thoroughly all over the place, especially around the fan. but you may be right, i've never done any of this type of stuff before, so i'm pretty much guessing on where to spray. i think my gpu is running a little cooler now (like 2-4c cooler). i didn't SEE any dust flying out anywhere though. maybe it's like microscopic dust? would that even make much of a heat problem? if there are specific areas that i should spray specifically, please let me know.. i was just guessing.
@forge, if you were replying to me, my gpu does downclock during idle. it is a nice and coolish 48c. after reading that post, that is quite interesting. do you have a link to a guide or something about applying paste on the gpu? i dont even know where it goes exactly... and i might not even be 100% where the gpu was when i opened the g51 up. so, that probably gives you an idea about my computer knowledge. -
CitizenPanda Notebook Deity NBR Reviewer
I would take it off and use some pads and isopropyl and rub off whatever you can get at, and then spray through the fins thoroughly.
Edit: if you can't take it off for warranty reasons, spray thoroughly through the exterior exhaust vent area into the machine. -
The dust in my 8800 GTX looked that way. I hadn't cleaned it in 2 years until a few weeks ago. The temps were hitting 100C+ during gaming so I figured it was time to open up the shuttle and clean everything out. After the removal of the dust and a nice application of MX-3, load temps dropped to 78C! lol.
-
alright thanks for the tips, i'll open it up again and try to find the heatsink fins. i know there is another panel (a smaller one) on the back of the g51... but i couldn't even unscrew it because the screws were so small, none of the screwdrivers fit. it's not located under there is it? and i dont even know where the warranty sticker is.. (not that it even matters now apparently, but i'd still like to be careful with it if i can)
-
CitizenPanda Notebook Deity NBR Reviewer
-
okay, yeah, i sprayed mostly around there. but the fan was there, i didn't see what your pic showed. do i remove the fan?
-
CitizenPanda Notebook Deity NBR Reviewer
Just blow compressed air into the fins. -
does it just rip the sticker? because as joker posted, the breaking of the sticker doesn't void the warranty, they just check more thoroughly that nothing was modded.
-
It doesn't break the asus warranty. Bestbuy does not have the asus warranty.
-
right, it has a bestbuy warranty.
-
oh dear, after about 3 hours of left 4 dead without more than a few seconds of minimizing, i reached a max of 99c. i need to figure something out now
-
Well you could take the bottom cover off and put it on top of a block of ice.
-
SoundOf1HandClapping Was once a Forge
Back with my old Dell, I did that actually. Took an icepack, wrapped it in towels, and put the laptop on top.
-
-
SoundOf1HandClapping Was once a Forge
Yep. No more downclocking when I played CS 1.6 for three hours straight.
-
sounds like a plan.
-
so, i've been trying a bunch of random things hoping for some magic. i blew compressed air out the vents as recommended with no result, and i've changed video drivers to 186.81 to see if the .82 was a problem for me. No changes in temp underload, but now my idle temps are 10c higher (i dont think the gpu underclocks anymore when idle). But I was having the nvlddmkm driver issue popping up every 10 minutes in Garry's Mod for some reason. Now I have no problem, so it looks like I'm sticking with these drivers for now.
So, it looks like I need to reapply the paste on the gpu. Something I know NOTHING about.I hear this arctic silver 5 is the way to go? (this it? http://www.newegg.com/Product/Produ...&cm_re=arctic_silver_5-_-35-100-007-_-Product )
I read a bit about applying paste and what not, but I'd like specific instructions and guidelines for the g51 since i have no accidental warranty and no experience with doing any of this type of stuff.. and I do not trust myself doing this without some sort of step-by-step guide for idiots. So, i dunno, does anyone know of a guide for this or anything? pics would help so much.
I feel bad for asking so much, but I really am absolutely clueless in this matter. -
SoundOf1HandClapping Was once a Forge
Don't use Artic Silver. G50/51/60 favorites are MX-2, MX-3, and ICD7. Also the Soviet Approved Shin Etsu X23 is a good one. MX-2 is probably the best balance of cost, ease of application, and performance. I use MX-3, though, and I'm very happy with it.
Anyway, just disassemble the heatsink assembly--which means the CPU, northbridge, and GPU interfaces,plus unplugging the fan. Four screws on the CPU, two on the northbridge, and two on the GPU. Remove.
Use a good alcohol cleaner to clean off the paste off the GPU, CPU, and heatsink. (Since you break the CPU contact you have to reapply on the CPU as well). I recommend ArtiClean, which is a good workhorse cleaner.
Make sure you don't touch the foam-looking thermal pads. Leave them be.
After cleansing the dies--and make sure they're squeaky clean, since any leftover paste will negatively impact performance--put a glob of paste on the GPU die. Use something thin and relatively rigid--putty knife, razor, etecetera--to spread the paste around. Since the heatsink isn't a perfect fit on the GPU, you'll use a thicker layer than what you used on the CPU. After completing that, repeat on the CPU, but this time put a paper-thin amount on the CPU, scraping off any excess.
(Optional: To test if you have a good fit, replace the GPU interface and press down where the GPU die is. Remove the heatsink and inspect it. If there's a nice, solid imprint of thermal paste, then you've done it right. Clean it off again and repeat, using the same amount that you used)
With that done, replace the heatsinks and screw down the interfaces. Plug the fan back in. Reassemble. Ta da. -
Dude I hate to say it, but I would just return it if you have the best buy warranty, if not send it to asus and tell THEM to fix their problem they created with this laptop, I had one of the first versions out of the store before they were officially sold! First time I ran crysis at full res high or med settings it got up to 95C+. I knew from the computers I have been building for the past 12 years that no laptop should have a gpu in it that gets that hot brand spankin new! So I reapplied as5 to it and slapped it together and the temps were even worse. So as quickly as I bought the darn thing I returned it! Asus does not know how to properly cool that gpu in their laptops nor any other system as you can see from the past with the older nvidia cards cooking in them.
FYI: Recently for the heck of it I got a good deal on a Toshiba A505-S6986 laptop. Specs are:
16" 1366x768 HD LED samsung panel
230m GT nvidia card with dedicated ddr3-1024mb ram
T6600 cpu at 2.2ghz with 2mb cache
4gigs of ddr3 1066 ram
400gb hard drive
windows 7 64 bit home premium
harmon kardon speakers which kick
crappy keyboard
nice touchpad
all the ports u could need
all this for 700 buxs
temps out of the box on gpu at idle 45-48c without fan going then fan kicks in slowly with no noise to keep it down at 45c
after playing crysis at 1366x768 all med settings at native res i get 30+ fps average and temps on the gpu card never get past 65C this with stock clocks
I way overclocked this GT 230M card past 240M GT clock speeds and am getting 7000+ points in 3dmark06 and 19000+ points in 3dmark03 and after playing any game for 30 minutes or more the gpu gets up to 65-70c with overclock settings and never past that temp, and this toshiba laptop only has one fan with side vent and side holes in bottom under it and fan at full blast is barely noticeable (all this compared to the asus G51 I had) and for a laptop this thing cools nicely with no cooling pad sitting on a wooden desk with 70F degrees in my home.
I think it blows the asus quality out of the water for a laptop. the only thing that sucks is the keyboard and the slower gpu but its not to shabby since i can overclock the hell out of it and temps stay low no matter what..
anyways back on subject... RETURN OR GET RID OF THE ASUS its one day gonna burn out on you and then what? to late to return or get warranty work, when I emailed asus about the overheat issues they said it was not suppose to do that and they wanted me to return it to them and wait months for it maybe to get fixed, I just returned it to the store and got full credit back!
Good luck! -
i've had it since july, i don't think i can return it now.
-
i already try to put new thermal paste ic7 on my G51VX-RX05 (BB version) i change the processor to X9100
are these temps normal? The screenshot is from 15 minutes idle, then running 3dmark 06 twice. can someone here tell me the normal temp for GPU??
and
can i replace the thermal pad in the northbridge to thermal paste without copper shim? -
i need to double-check, does applying thermal paste on the gpu and cpu void the warranty? if so, i'm just gonna backup my stuff, and wait for this baby to burn (hopefully within the next 7 months)
-
Soviet Sunrise Notebook Prophet
-
,
should i take the 1mm or 1.5mm of copper shim?
something like this http://cgi.ebay.com/HP-TX1200-TX130...iewItemQQptZLaptops_Nov05?hash=item3ca56d1a8c
is my GPU and CPU temps normal at these rate, or i have to reapply ic7 on cpu and gpu? -
Soviet Sunrise Notebook Prophet
I've never gauged a G51 before so you will need to find out for yourself. Because the northbridge shares the same heatsink with the CPU, you need to get the thickness right or else there will be insufficient contact with the CPU or the northbridge. On my M570TU, my shim was about 1.1mm when measured with my calipers. So I would say get something around 1.25mm and lap it down a bit. Using a copper shim will help you with FSB overclocking as it will keep the northbridge cool.
-
SoundOf1HandClapping Was once a Forge
The shim I inserted is .55mm, and as far as I can tell during my times repasting, it makes full contact.
-
Soviet Sunrise Notebook Prophet
You heard the man. It looks like somewhere around half a millimeter is the way to go.
-
, what about my temps is it normal, so tell me the max temps for the GPU and CPU?
-
thanks all
-
93C is not bad, atleast its not on fire. Remember furmark burns all your gpus hardware facilities.
Have you also tried OCCT? -
SoundOf1HandClapping Was once a Forge
OCCT GPU test gives me higher temperatures than "the usual" Furmark settings.
-
CitizenPanda Notebook Deity NBR Reviewer
-
-
CitizenPanda Notebook Deity NBR Reviewer
-
-
CitizenPanda Notebook Deity NBR Reviewer
My Chipset idles at 45C, the CPU idles at 28-29C @ 3.5. The Chipset loads at around 75C playing games. Intel Burntest does kill it though (hit 89C before I stopped the program).
What Chipset temps do you get? -
-
SoundOf1HandClapping Was once a Forge
I maxed 75 for the northbridge for about seven minutes on the OCCT power supply test. I stopped it after my GPU went over 96C, though.
My G51vx's gpu is getting hotter and hotter. What do i do?
Discussion in 'Asus' started by Twistamafoo, Nov 10, 2009.