My G51 is a hot potato,Help!

MX-2/3 don't require a curing period, either.

Well, I'll give a rundown once I get it. Fifteen bucks with tax and shipping.

I have a set, I was just wondering if it'll work just fine. Apparently, yes.

I'm not quite sold on Diamond 7, but who knows. With a (notriously) hot-running laptop like the G51, every degree counts.

Plus it's gallium. From what I've heard from other people, it still tends to bond with other metals.

 

Everyone in the desktop world that uses water cooling considers Petra's Tech Shop as one of the best stores for specialty water cooling and air cooling parts and equipment. They also carry their own line of water cooling kits that cover everything from entry level to state of the art. I don't buy from them often because they sell almost only desktop parts, and the times that I do buy from them are for bulk orders for my friends and clients.

The good thing about them is that they are in San Jose, which makes USPS FCM within a 1-2 day shipping range for us. Also, they handle each order with great care. They sign a personalized note on each packing slip and include "Petra's pen" as a free gift. I have yet to hear someone who has had a bad experience with them.

 

My reseller version of the A1 seems to be OK. With CPU running in Asus supplied Turbo mode (2145MHZ) and GPU overclocked to stock settings I'm running about 50-51C in idle and 70-80C running games, surfing and watching movies.

I've also clocked GPU up to GTX 280M settings and ran for a brief period of time. Heat did run up to 80-90C.
Has anyone seen the operating specifications for this GPU chipset? I did see where the Nvidia GTX 260 was rated up to 105C as a maximum. Even at these temperatures the chassis remained cool to the touch.
I've also got the larger 9 cell battery on the back. Took about 2 hours to completely charge while running the pc at the same time. No real additional heat noticed.

http://www.nvidia.com/object/product_geforce_gtx_260_us.html

Something else to keep in mind. Asus provides some overclocking apps on the PC and provides a 2 year warranty. They must have faith in their design otherwise the standard warranty would be like everyone else at one year.

 

The GTX 260M's critical temperature is actually 108C. After that, it will automatically downclock to prevent hardware damage

 

Once I change my CPU, and apply it on the GPU, i will not remove heat sink anymore... so I can "suffer" the moment of application, to get the better performance in the end
But, when you use a small brush, its much easier. I applied this paste once, and its a whole different procedure than any other paste (but harder, and more time consuming).

I'm thinking is it worth to apply it to the G51?
The heatsink is not alumiun right? But what is the CPU core made out of? Is there any danger it will react with the heatsink/ CPU?

 

Why did you link to a desktop GTX 260? Regardless of the maximum rated core temperature, nobody should be pushing a GPU that high in the first place.

108*C is for the core temperature only. GPU memory has a much lower temperature tolerance generally in the low 90's, does not have temperature sensors, and does not have an auto downclock threshold like the GPU core. Forgive me for being such a conservative overclocker, but if the entire Asus forum wants to continue to shake hands with the devil by dancing in the 90's and above, rest assured you all will be seeing symptoms of a dying GPU in the future, and even more so if the temperature rises past the high 90's and 100's. Don't ever forget that the GPU memory and the GPU core share the same heatsink. Does that mean that you all should keep your GTX 260M's at factory 500/1250/800 or similar? Soviet says yes if it keep the temps out of the red zone. But that shouldn't stop anyone who is looking for a little bit more kick out of their GPU. My best recommendation is to push the clocks to 550/1375/800 and lock it there. I'm assuming many of you G51 users have 1366x768 screens. 1366x768 doesn't need that much memory bandwidth and you barely see any FPS increase from cranking the memory, and memory produces a substantial amount of heat. Lastly, I'll leave you all with this. True overclockers are concerned about cooling, not performance. It makes me cringe to see other users casually push their systems into such temperatures.

http://forum.notebookreview.com/showpost.php?p=4875854&postcount=117
http://forum.notebookreview.com/showpost.php?p=5175428&postcount=41

I suppose you can have your GPU linger into the low 90's occassionally if you plan to buy a new computer within the next two or three years. But for some of us who value machines just as much as life, I would never abuse such a computer that performs all the tasks we ask it to do without question, stores our most prized sentimental memories, connects us to the world, and depends on our love and care, like that, ever.

Using a brush to apply thermal paste will just waste paste. Tear off a piece of pressed cardboard or find your old credit cards and use those. It really shouldn't take any more than 10 minutes to apply paste on the CPU and GPU.

You can follow this short visual guide as reference. http://forum.notebookreview.com/showpost.php?p=5027711&postcount=520

If you mean ICD7, then yes. Just make sure you get it right the first time so that user error doesn't skew the performance of the paste. Don't even look at Coollaboratory Liquid Pro. You'll thank me later.

Your CPU heatsink is copper, and the black CPU die is acryllic.

 

There is also no way to tell the temps of the memory on a laptop video card. IDK if you can on a desktop since everytime I head to building one...I end up with a laptop .

Once our gpus start to stay within the 48-95C range , we should be fine.

 

About a year ago, my buddy and I removed the backplate of his M860TU and subjected his machine under the temp gun. After running Furmark for a good amount of time, we gauged each part of his copper encased overclocked 9800M GTS. The heatpipe over the GPU core was 85*C, the copper wings that cover the GPU memory was 82*C, and the MOSFET's were hovering around 78-80*C. Keep in mind that the M860TU has a more robust heatsink.

 

I never asked to prove your statement

I said there is no sensor for the memory chips on the gpu. I don't think were all going to remove the bottom panel while gaming and point a laser temp reader at the gpu.

 

The point that I'm trying to make is that we should, as much as possible, avoid pushing our cards that high. My support is that the GPU memory will run very close to the core in terms on temperature, though this may vary as my test was done on a different system.

 

Thx for the advice however I was specifically talking about the Liquid Pro which is nearly impossible to do with a credit card, as its consistency is basically liquid.
I also thought the paste was dangerous, so my final decision is IC Diamond 7.

I agree with you. The system runs hot, and therefore we cannot increase the heat (which decreases its life). We mod and strive for lowest temps or not OC beyond some point.

This review says otherwise.
http://aphnetworks.com/reviews/arctic_cooling_mx_2

- All pastes were given an adequate amount of time to settle. (Break time - Arctic Silver 5 and Arctic Cooling MX-2 both has an approximate break time of 200 hours)

Maybe thats y some people see an initial increae in temps and a slow degradation after some time.

 

Yes, Liquid Pro is very dangerous as a thermal interface. I see it to be more of a poison because you can keep the syringe in your pocket and stab someone in the back of the neck with it. I wouldn't want that stuff in my bloodstream.

 

I gave up this , too much trouble to go through, I'll just use it without payin attention to the temp n if it dies in 2 years, I got the warranty, beyond 2 years, I'd buy a new laptop that's it.

 

The memory chips should run cooler since they do all have thermal pads on them. The reason most of us and manufacturers use thermal compound on the core is because the gpu core was mainly designed for that(metal interface) and that the copper plate needs to be almost dead flat on the core so that the memory chips can also touch the rest of the heatsink,

 

It really surprises me how some people believe that thermal pastes such as MX-2 doesn't require break in time just because the manufacturer says it doesn't. All thermal pastes need some degree of break in time even if the change in thermal performance is relatively minor. Is the break in time as long as 200 hours? Not really. But it certainly isn't going to be as short as the first few hours after application either. What makes MX-2 unique is that it is primarily made of aluminium-oxide rather than silicone or metal pastes, or cermamic pastes, which suffer from short shelf life.

 

The memory still shares a lot of heat with the the core. It is the same principle as to why an idle CPU heats up when the GPU is stressed since they share the same heatsink. However, on most desktop cards, the memory heatsinks are independant from the core and do not require thermal pads as memory heatsinks do not put potentially damaging pressure onto the chips as opposed to a large one piece heatsink tightly clamped down with screws.

http://forum.notebookreview.com/showpost.php?p=4965675&postcount=133

 

I talked to ASUS support and they told me I could replace my thermal paste and NOT void the warranty even though I would have to rip the sticker off. He said that as long as there is no user damage then the warranty will still be valid. I have a GPU that gets up to 102C as well and I replaced the thermal paste on it and my CPU and it only helped for a couple of weeks then the temps went back up. As of now the laptop works fine and I have never crashed or artifacted and I cannot be bothered to RMA it as it would take weeks. I would just say use your laptop if it works fine and forget about it. You have a 2 year warranty so deal with it later!!!

 

Have you guys thought that maybe it's a problem with early adopters?

I know my first G51 hit 102C in Crysis with stock clocks (and had artifacts while doing it). Furcube is far more stressful for the GPU and I've been able to stay around or under 97C at 1280x1024 and 16xMSAA while overclocked. The second one I received was a little over a month after I got my first one. Idle temps are also under 50C- they hit 45C at work (room temperature was 74F).

 

Random question but do the BB G51s have a 12month asus warranty?

 

Not unless you buy best buys one. Forgot how much it is, pretty useless in this case since they have a 2 year warranty.

 

I said BB G51 meaning bestbuy hehe. I should have spelt out bestbuy my bad but I didn't understand your response.

Do the bestbuy G51s have a 12month warranty through asus? I'm just wondering because the bottom panel has the sticker that has the model, serial number and it says Warranty: 12M

My G50 also had this.

 

To my knowledge, all Asus notebooks purchased via BestBuy have 12mo warranty service provided through BB/Geek Squad, not Asus NA in Fremont.

 

So it was 3am and I decided to finally try the gophn focused airflow mod.
Heres the video to get you up to date on what that is:
http://www.youtube.com/watch?v=QVWk2ywIsWw

Remember I also have the mod which I did to the bottom panel, I cut out a hole where the cpu/gpu fan is located.
http://s52.photobucket.com/albums/g33/motok155/?action=view&current=IMG_0299-2.jpg

I placed window netting as a fliter for the fan(update once I get a chance to take a new picture)

I took an old box from my old bindings and cut it out in a rectangle to block the nc-2000s vent completely. I cut 2 holes. One in a circular shape to go right beneath the hole I mad in the bottom panel so air goes up into the fan directly and another rectangle like shaped hole beneath the vent for the gpu so air rushes up to the gpu.

So far my high temps in crysis wars dropped from 94C to 88C. My house being 76C. My gpu idles 44-45C and cpu at 29-30C

Once the estore has some more bottom panels I plan on buying some 4000 rpm 80mm case fans and making my own cooler and placing a fan where each remote vent is, rushing the air through the components up to the fan.

 

My god, man!

 

hahaha, has your skin melted off your bones yet?? ;P

 

sure they do...most asus laptops are covered. Except yours, since you changed your cpu

 

o my bad I meant 76F sorry my fault haha.

Anyways GTA4 highest peak is 84C with the mods. Thats a good 10C drop from before the rush airflow mod.

 

Hmm. The AS5 isn't cooling down much. My GPU temperatures are still worse than they were with stock paste.

Munchkin, when you spread your paste, was it real thin? Mine had to be significantly thicker than what I put on the CPU.

 

The gpu wasnt too thin, I might redo it. My cpu is very thin.

 

When you apply thermal paste, NEVER spread them out manually. A single small blob in the middle is what's needed. As you place the heat sink on to the GPU/CPU die, the paste will spread itself.

Forge: applying too much AS5 will often increase your temps. Also keep in mind that there is a curing time (~150 hours) before AS5 becomes fully effective. ICD7 on the other hand, does not have a curing time.

 

It has about 2 hours, but thats nothing compared to the AS5.

 

True, 2 hours is what they've listed, however I was able to reach peak performance almost immediately (within 30min)

 

As5 is 200 hours, I posted it earlier on.

 

MX-3 works right from the get-go. If it does have a slight curing time, then I'll be even happier. MX-3 slaps AS5, steals its wallet, and leaves it unconcious on the street.

Under identical testing conditions, stock paste gave me 91C on a 3DMark trial, freshly applied AS5 94C, and MX-3 88C.

 

So in order of preference it'd be

ICD7>MX-3>AS5 ?

ICD7 is cheaper than MX-3 on ebay so might grab some. I've read that it's really hard to apply though...

 

Haven't tried ICD7, but MX-3 was ridiculously easy to apply, all I did was run a line down the middle of the GPU and let the heatsink spread it as needed as I set it into place and screwed it down... and right off the bat, temperatures were noticiably better.

 

Ok, I am having some troubles with my G51vx-rx05 -- upgraded to T9900

I posted in another topic and I believe that Soviet helped me out, I had removed my thermal pads from the gpu memory, and replaced them with OCZ freeze paste. The reason was the pads were falling off due to removing my cpu and replacing it twice with newer.

I ordered new 1mm thermal pads from a link soviet provided, and today cut and placed them very neatly in the correct spots. I them re-applied new thermal paste (OCZ Freeze) to the gpu die, cpu die, and northbridge die.

I ran L4D, and I immediately noticed my fans kicked up to high and I alt-tabbed out in game to see temps in HWM and saw 105C max temp, and around 100C current on the GPU. CPU temp was 56C max, and 50C current.

I applied the thermal paste in the middle of the die, about a small grain of rice size. I did not spread it out, I just press the heatsink down on it to get even coverage. Is it possible that I'm applying too much or too little? Should I spread the thermal paste out with a credit card as others have suggested?

I am using 186.82, and am going to go try 195.62 WHQL. I will report back with readings in a little while. Any help would be greatly appreciated. Thanks!

GPU Idle temp right now is reading 54C, and CPU 34C.

 

Anyone?

Tried 195.62 and am now idling at 79C, obviously not down clocking, another user reported it was working fine. Has anyone tried the dox version of the drivers and noted a difference in temps?

 

to get it to downclock google the powermizer switch. and I would definitely spread it out first, the heatsink on the 260m doesn't put enough pressure to really spread it, let alone spread it evenly.

 

Seeing this a lot in this Forum, you guys do realize that you can install Evga Precision and set it up so tht you can fully monitor GPU Temps, Core Speeds and FPS while "in game".....With the 260M you won't be able to utilize Precision to OC your GPU, but you can monitor while gaming, extremely useful tool.
Yes, 195.62 WHQL solved my heat n downclocking woes, remember, not everyone will have te sam software config as mine and this likely has a lot to do with how 195.62 performs.
Also, these are Notebooks, meant to run hot and not meant for treating like you would your custom built performance desktop, as in OC'ing GPU's etc etc......
My desktops in my Sig are performance Rigs which I OC, run benchmarks etc etc.(Vantage P34.5K on the I7 Rig)....certainly not my Notebook, thier simply not designed for this. As far as Vantage / 3D Mk scores..could care less what my Notebook yields, provided it stands up to some casual gaming...which it does.

Did you do a clean install of the 195.62 WHQL ?

1. Remove PhysX
2. Remove Driver
3. Run Driver Sweeper and set to remove all nVidia Driver / PhysX components
4. Go to C:/Windows/Prefetch and delete all Nvida related items (nvxxxx files) many of these aren't found even with Driver Sweeper.

Then install your new Drivers.

 

ok I just played another game, after tearing it down again and reapplying the paste. I spread it evenly over the die with a plastic card. It read a max temp of 107C after only 1min, and around 100C current! Though the game was running fine? Also my cpu only hit a max temp of 59C, and was around 52C in game.

I am really confused and looking for help from anyone. I wish I had never open up the gpu and put on new paste. This has been nothing but trouble.

Should I try using a thermal pad? When I applied new paste again I noticed it had not even spread out on the die, and I assume maybe the copper heatsink isn't making full contact to the die. Can you use thermal pads on gpu dies?

Thanks again guys.

 

^^^^ Is the Warranty not now Void after all this tearing apart and re-applying TIM?

Personally I wouldn't go thier with a Notebook under Warranty, different ball-game all together with a Desktop.

With that said, a small dab of TIM in th middle should do (let it spread with its own pressure), never spread it out evenly, too much TIM will be just as bad, if not worse, than too little.

 

Sorry, just saw this thread, been gone a couple days.

The problem you are having is the same problem I had with AS5... there is a slight gap between the GPU die and the heatsink, and mine was even warped a bit, ie didn't sit perfectly flat. I could tell this, as like you said, when removing the heatsink, you can tell that it didn't make full contact with the heatsink/GPU.

What I had to do was put a thicker coat, much thicker than I normally would of AS5 to get it to make contact. This is not ideal, but did lower temperates to back at stock levels.

I'm not sure about the paste you are using, but I ended up with MX-3, a non conductive paste. What I did was run a line down the length of the GPU die, and then seated the heatsink and screwed it down, and have been running great ever since, cooler than stock, and very happy. The MX-3 is nonconductive, so I didn't have to worry about it spreading out a bit and possibly touching something and shorting it out, and flows really well, so the heatsink spread it properly and made it's own contact patch as needed. This is how MX-3 instructions recommend using it.

Your paste may be very similar, not sure, but reguardless, are very thin application is not going to work on the G51.

 

OK, just checked your paste, it seems it is not conductive, so I'd recommend running a bead left to right on the GPU (in relation to the laptop itself), and let the heatsink spread it.... if yours is anything like mine, a grain of rice size won't do it, and it'll just make very partial contact to the heatsink/gpu. Worth a shot, and like I said, even with this technique, mine is running quite a bit cooler than stock, so very happy.

Definitely don't game with it hitting those temps though.

 

Thanks for the response! I thought I figured it out lastnight. I removed a few thermal pads that weren't on any of the gpu ram because like you said my heatsink isn't fully contacting the gpu die. This made it contact more evenly, and I went to reapply the paste, but ran out. I had an extra tube that came with my T9900, it's Shin-Etsu MicroSi. I placed two large lines on both sides of the die, and pressed the heatsink down. When I removed it to see the coverage I saw that it had covered most of the die. When I ran L4D it immediately was cooler, at 80C for a few min, and then ran up to max temp of 95C after 15min of play.

But today I again hit 102C max temp, and it was in the 90s during play. I guess this is a lot better than what I was getting before, but still I'm concerned. I called a local shop that builds customs desktops and does laptop mods and the laptop tech told me that the TJ Max for the 260GTX M may be really high, and that those temps may be perfectly normal. Does anyone know the max temp for this GPU? I haven't had any shutdowns or artifacts yet.

 

If temperature's over 100C you should be worried.
when I had g51j my temperature wasn't that bad, max temp from playing Dragon Ages was around 80C.

 

I have a g51vx x1a and my gpu usually tops at around 98C.

 

Crysis, MW2 etc Max "in game" 90c

 

There's going to be so many fried G51 260Ms in the next few months. The max temp is about 110C for these things. Hitting in the 100s consistently is going to accelerate the wear on the GPU as well as the other components in your notebooks due to heat build up.

 

Yeah, but I'm planning on selling it soon and its under warranty so meh.