G73JH - is a gpu temp of 98°C acceptable?

i ran a 1 hour session of furmark yesterday, at first the temps quickly reached 95°C, then it took about 8-10 minutes for each additional degree until it reached 98°C. it stayed at 98° for about 30 minutes, so this seems to be the max temperature of my gpu. is this still acceptable? usually it´s said that anything below 95° is acceptable and anything above 100° is not. and that such a high temperature will reduce the hardware´s durability.

so is my system faulty or is this still in the range considered acceptable? i really dont want to change the thermal paste myself unless its really necessary. .....

 

What's your ambient? That's a bit high, but I think tolerable. My computer tops at 84C after running a few Unigine benches, but my ambient is 74F.

 

the room is at about 24-25C... so this should be no excuse...

 

I've been running Furmark on extreme burning 1920 x 1080 no msaa for the past 2 days and my gpu peaks at 98*C but holds 97*C for the most part.

 

That's on the warm side but it's tolerable.

 

Not the best, but tolerable for furmark on these machines i guess...

 

Those temps are fine. You will see 90-110C on these notebooks with Furmark, depending on ambient temp and humidity. At around 110-115C it will shut itself off because that's the thermal limit they specified.

Well I suppose if you sit outside during the winter that it might go lower. Assuming you have a real winter wherever you are.

 

jebus does this laptop reach this kind of temp when gaming, lets say bc2? Because i was thinkin of buying this laptop and after looking at this post its put me off a bit

 

Users have reported a wide range of temperatures while gaming, anywhere from the low 80s to shutting down after blasting through 110C within minutes. It mostly depends on how well the thermal paste was installed at the factory (and therefore your luck).

 

In the interest of compiling data from different sources:

High res [1920 x 1080] FFXIV benchmark my GFX temps peak at. . 98c.

In regular games, it peaks usually in the mid 80s [according to HWMonitor]

TBH, it startled me just a little bit seeing the same number pop up on a forum on the subject

 

It depends a lot on the game or "benchmark". Furmark will get it hotter than any game because it's very artificial. Some games will not heat the chip up much at all while others will get it near Furmark temps.

Don't worry about the temps. It is designed for them. Only worry if you get a lot of shutdowns or screen corruption because that means that the hardware wasn't assembled properly or it is defective. That's when you call upon your warranty.

And btw all of these high-end notebooks run this hot when gaming.

 

The card will deteriorate running at such high temperatures every day. Its certainly not true high end notebooks run this hot.
None ive had before ever came close to these insane temperatures (all had the best gpus of their time - crossfire 3870, gtx 280m, mobility 5870). And ive even had the same GPU in the previous one (Clevo D900F) which didnt even get near 80C in anything and even a crossfire 3870 (with even a shared heatsink) one didnt heat to such high temperatures.
The shutdown limit (design to prevent damage to GPUs) on the g73jh is 105C on older bios and they set it to 110C on 209 to get less rmas due to shutting down. On the D900F the shutdown limit was at 100C. That tells you something.

 

Well, that concerns me a bit then, as I just got done running Furmark with the stock drivers, high-performance mode, 71f ambient temps, 1920x1080, xtreme burn, post fx, 8x msaa and ended up with max temps of 106-107c after an hour...

I coudl swear I tested this machine when I first got it and it only got to 96-97c at most.

I highly doubt the heatsink fins are clogged, but will blow it out later today to just verify it...

Is the TIM used on the GPU just that bad? I do have some Tuniq TX-2 on hand I could redo the TIM with, but I want to make sure it isn't worse than what they used at the factory.

I really like this laptop, especially for the performance, but the damn keyboard missing characters and now this is really beginning to me off for a $1800 purchase. Hell, I already had to take the unit apart enough to push on the cabling for the ethernet port because that didn't work from the factory.

I will say, my son gamed on my laptop for about 5 hours yesterday playing Starcraft 2 on Ultra settings without a single issue. I just don't want there to be an issue on the next 6 months. I really hate it when equipment has issues when you need it the most.

 

"i ran a 1 hour session of furmark yesterday, at first the temps quickly reached 95°C"

Are we talking about TS0 or TS1 temp? As for TS0 (DispIO) my highest was 84C I believe, with an ambient of 28C playing Crysis. As for the TS1 (MemIO) (same time, same conditions) I got 97C.

Could somebody clarify it to me which temp is the discussion about?

 

@ mrwhiteshadow
Were talking about the highest temp - the memory IO (TS1).
Your gpu is getting pretty toasty too...

@The_Beast
Its not your TIM, its the lesser quality parts Asus used to make their GPUs, so they run hotter than the competitions GPUs. It could be that the temperatures are higher now is becouse your ambient temperature is higher (its hot as hell here this summer). Or perhaps the clocks changes in drivers affect your GPU this way. In any case, the higher temperatures are the hardwares fault - they should never reach such high numbers in normal conditions.

 

HellCry you don't know what you're saying. Just do some searching about notebook GPU temps. Look up GTX 280M or Mobility 4870. I had a GeForce Go 7800 GTX in my previous Dell note that I could break 100C with in Furmark and I had it ran up until I sold it this spring (that's about 4 years).

Some desktop cards run 90+C. I know of some 4850s that ran that hot or hotter because ATI wanted to keep the fans quiet and the engineers said the chips could take it. I don't know how long you've been gaming but you can go all the way back to Voodoo3 and find 90C because they didn't put a fan on them. People were whining about that to no end back in 1999. I still have a pair of Voodoo3 cards that work fine for the occasional retro gaming.

Get over the temps. They aren't going to kill anything. The GPUs have various types of thermal protection that will prevent overheating damage. If you get a notebook that's flaky from the start, it's not because of heat, but due to some sort of other defect. Hopefully the companies have mastered RoHS solder now too and we don't have failures like NVIDIA had a couple of years ago.

Of course I would love it if we could have huge GPUs that ran at room temp. I'm sick to death of hot computers. But it's pretty clear that you don't get fast GPUs that push the limits of manufacturing and get a cool chip. And in notebooks where space is ultra tight, they really have to push the limits. But have you noticed how a Mobility 5870 isn't based on the Cypress 5870 chip but instead based on the 57x0 GPU? They knew the limits.

 

HellCry is right. The temperature of the G73JH is unacceptable hot. There is no way no how Ati designed the TSS1 temperature to be 15 degrees and more hotter than TSS0. It is a badly placed and badly pasted heatsink that is causing this.

After repasting and properly placing the heatsink my TSS1 has a 6 degree difference which is the same for Alienware laptops( see other temperature thread). My temps are exactly the same now as an Alienware 15mx after a long gaming session of Bad Company 2. http://forum.notebookreview.com/alienware-m15x/474717-m15x-ati-5870-mod-82.html

The heatsink is also not of a very great quality either. This could haven been done better which could have helped somewhat more.

The other problem is that the themal shutdown temperature is not measured on TSS1. This explains why for some people the laptop shutsdown at 102 degrees while it for other people doesn't even shutdown at 109 degrees on TSS1. TSS0 is used for that and with the bad pasting job it is always significantly lower. Or in other words. The laptop shutsdown too late!

90 degrees is indeed fine when gaming but with the bad job Asus has done TSS1 is at that point 105 degrees or higher and that is not good.

 

I guess you are in close contact with an anonymous ATI engineer who's throwing up red flags over the impending doom of these notebooks then right?

I wouldn't mind having another heatpipe though.

 

Youre talking about one of the bigger blunders ATI made with those passively cooled 4850 cards. They ran hot, some died on the users, many users compained and some sought huge aftermarket passive coolers to cool them down. I dont see why youd bring that up or how it would illustrate your point? Manufacturers make mistakes as well.

My friend had a M17 which had a bad TIM application and he ran in mid 90s, and one of the cards gave out on him just under a year (the master card). I reaplied the paste on mine to make it run in the high 70s and the new owner is still using it normally.
Heat certainly isnt a friend of gpus or other laptop components any way you put it. And were not talking about huge power hungry gpus in laptops (such as the fermi desktop cards, which in my oppinion are a huge fail becouse of the enormous power consumption and heat they give out), but small, low power solutions which certainly enough can run cooler.
If you had a hot running gpu, either becouse of a bad TIM application, poor cooling or just a poor card, i feel sorry for you, but its not only possible, but very common, for gpus in high end laptops to run at normal and acceptible temperatures and thats what i expect, since i dont want the thing to die on me either within the warranty period or after. Also dont appreciate the heat ouput and noise when it heats up.

I can compare apples to apples. The fact is, if you compare systems that have the TIM applied properly, the Clevo/Dell/MSI Mobility 5870 run a LOT cooler, period. I had one in my D900F so i can tell that first hand. That card never touched 80C even on furmark in that laptop.
It can run cooler, but the Asus GPU is lower quality, as well as the cooling isnt full copper so it runs much hotter. I just expected more from "silent bomber" and "cool under load" advertisment.

 

I'm not gonna argue with anecdotal evidence. We'll see how long these machines last. Feel free to dig up an ATI engineer and have an anonymous conversation with him about what he thinks.

What we have more than ever before is defective hardware. The GPU corruption stuff is amazing to me. These products have never been as flaky as they are today. This kind of thing makes people wrongly worry about temperature when there are clearly other factors at hand. There are too many variables to just jump on the obvious "it's too hot!!!" thing.

But I'll tell ya that since the 486 people have been all worried about temps. A lot of forumgoers who know not what and a lot of engineers who don't want to get involved.

And actually this G73 does run considerably cooler on the outside than my Dell with the 7800 Go GTX did. I can game with the G73 on my lap. That was not an option with the Dell. So I am quite pleased with that anyway.

 

I am talking to one, as well as a tech here .

But we do agree its the cheap hardware Asus chose that is faulty as ive been saying, i already said its not the heat responsible for GSODs, but hardware.
Although the temperatures do worry me and i personally think that an expensive high end machine should not run this hot, especially since i know it CAN run cooler, but this is not my priority right now. And the outside is pretty cool i do admit but id rather have a hotter chassis and lower temperatures on the card...
Lets leave the temperatures aside for now, whats more important is the hardware/drivers incompatibility problems were having with these Asus gpus...

 

After modifications my G73JH is fine so far. I uninstalled a load of software and repasted the GPU which shouldn't be necessary because Asus should have checked all of this already.

If Asus doesn't manage to bring an updated video BIOS and fix other problems with this laptop I would say my next one isn't going to be an Asus. To bad Dell is so bloody expensive with theirs and they borked the design as well now.

 

Look at Clevo. Perhaps that one will suite your taste if you like plain notebooks like me.

I know im never buying an Asus again, should have just stuck with Clevo this time as well...

 

If you are suffering from gsod's, can you rma it and get a new gfx card put in under warranty?

 

I wont rma it and wait 2 months to get it back, since it wont fix anything. This is actually already the second card i have with the same problem and the tech im talking with tested 4 random units for me in their repair center and they all got GSODs. I already told you its all the Asus GPUs that are faulty.

 

Apparently Asus didn't follow reference design, nor proper quality control specs for the device. Mediocre paste, (wait for my pics) pasting, and not tightening the gpus enough can all be contributing to the performance issues.

As for temps, well going from GSODs in 2 minutes with 101C+ temps in tss1, high 90s in the others to 2hr+ runs with 80c temps in tss1 , 78c in others speak enough about what proper pasting with quality paste and heatsink mounting can accomplish.

 

I don't know. I've had my G73JH-A1 purchased from Fry's Electronics at the end of May. I've had it overclocked at 800/1100 the whole time and have only had two crashes since then while gaming on BFBC2, and nothing else.

My Temps usually hover right where the OP's is. Around 95-97F.

And the thing never crashes. Ever.

I just played Metro 2033, arguably one of the most taxing games on the market for GPU at Normal
settings on DX11 for 6 hours, running it HDMI to a 60" Pioneer Kuro. The temps stayed at 94-97 the whole time, and not one crash.

Now admittedly I have a GPU that seems to have the Hardware I'D of being the one that isn't causing issues. I can't remember the thread where this was discussed, but I do remember Chastity posting about it so maybe she has the link or can provide a better frame of refrence as to what I'm talking about.

I believe there is a lot of misplaced and unneeded fear on these temps. I would be checking the chip I'D on the GPU and if it's not the one known to be stable, then RMA it to Asus for a replacement.

I can't comment on the buildmanship, but as for putting Dell on a pedestal above Asus, I wouldn't. I just also bought an MX11-R2 and there are build issues with the hinges on the panel to my own having dust under the screen requiring an in home panel replacement from a Dell tech. Not to mention that the picture quality on said panel is half as good as my G73JH-A1 even though after taxes, the Alienware still ended up costing over $1500 even though the A1 does double the performance and has way more features. Not to mention that Asus doesn't try to control driver updates and force you to go through them to update GPU drivers. Only to have to rely on smart forum members here to write INF files because the brand company (Dell) controlling your software is 3 driver updates behind the GPU maker. That alone will keep me from ever buying Dell again (Though I do truly love and recommend the M11x the same as I recommend the G73)

So there's issues with every make and manufacturer.

Instead of going through the fear or worry or hastle of repasting your GPU, go about your life and enjoy your netbook. If you buy a G73 it comes with a 2 year warranty already yes? Well, if the unit breaks down, use your warranty. Yeah it may be a pain in the , but I wager if you just game on the machine, if it did break down at all, it'll happen far enough at a point that replacement units will be plenty available. If that's not enough most brick and mortar and online offer extended product replacement gaurantee warranty's. I bought one at Fry's and I know if my unit dies I take it to them and they give me a new unit right away. That plus Asus premimum warranty that comes with the G73 is enough for me.

And before I got sidetracked about Dell when talking quality and buildmanship, I was going to say, while sloppy applying of thermal paste is unacceptable, there shouldn't be any surprise at the quality of components or thermals
being used. This is a machine that offers specs and components that cost nearly $1000 more at least from competitor's like Alienware and even more like $2000 if you want to get into a Cleo or Malibal. I paid $1650 for my G73, I don't know about you.

In China where most all of these laptops are made, Asus is considered high quality and a premium brand.

So my advice is just use your computer and game. It comes with a 2 year warranty. How many gamers stay on the same machine much after 2-3 years anyways?

If it runs at 98, just let it. As for the argument that these temps will slowly degrade the Hardware to a point of complete failure, I'd submit that the machine is performing within spec, and that actually those that insist on continuously running futuremark to test temps are creating that reality much faster than those that are just gaming on their machine's at 98F 24/7

I agree that if you're continuously having problems with GSOD's check the chip I'D and try to get it replaced with the stable one.

If you already have that and have no stability issues and are just tripping on your "gaming"
notebook hitting 95-98, then I say you're just creating unnecessary fear and worry and are shooting your self in the leg on enjoying what is one of the most fantastic laptops and gaming machines made thus far.

Tripping about 98F when your machine's running stable with a built in 2 year warranty is a fool's dance with fear.

You guys go ahead. I'll be over here gaming on my A1 at 97F for hours at a time, worry free, having fun.

Cheers!

 

^^ Excellent advice ...

 

Don't forget, if you aren't suffering any dangerous heats. RMA it at the very end of your warranty. g73 probably won't be in the production line anymore, so Asus will be forced to reimburse you with their next lineup.

I did the same thing with my M1710 and M1730.

 

Where can i find these Device IDs? I searched and googled and came up emptyhanded. Can you post the link or IDs please?

 

I can't find the thread either, but I know it was a thread that had something to do with checking your "Driver Packaging Version" and/or "Catalyst Control Center version" under system information. The thread stated that if one had a certain match on one of those numbers then you probably had a non faulty GPU. This might be in relation to the vbios and how everyone flashed theirs. I'm not sure, I haven't been around here for a while. I never did touch my Vbios though.

Again, if Chastity sees this maybe she can add some insight as I remember she was in the thread. Apologies if I'm throwing outdated info. I just distinctly remember a thread that said if your driver packing version or Catalyst Control Center version matched a certain set of numbers than you probably had a stable GPU. Mine did, and like the thread said, I've had no problems.

I really apologize for not having more info. Maybe Chastity or another poster will know what I mean and provide a link.

Cheers

 

As time goes on and fans degrade etc and the temps are really bad 2.5-3 years in (I got an extended 3rd year warranty; its 90 bucks but worth every cent if anything breaks down!) and I start getting more gsods etc, would they opt for replacing the GPU/motherboard etc or just replace the model? I know Dell has a VERY large overstock of 5 year old model motherboards etc for replacements haha

 

All you can find in CCC are driver related version numbers, no device IDs, so im guessing youre refering to what we already know but arent satisfied with - the old 9.12 or 10.1 drivers work stable, and newer ones dont.

 

sorry, but I never posted about the gpu hardware IDs

 

Sorry, then that's probably what it is.

But I disagree that the newer ones don't work.

I've not had one crash with 10.5, 10.6, or 10.7

I'm running the unit with the creative drivers installed and keep it constantly clocked at 800/1100 (I tried other's method of 705/1100 but when my system showed the same temps at those clocks as 800/1100 I switched back)

Every software that was installed with the machine from the factory is there with no conflicts.

The one thing I haven't done is try to constantly run futuremark. I ran 3DMark06 once or twice to see how the machine performed and that was it.

I'm sure if I ran that BS program long enough my machine would overheat and shut down too.

But I don't. I just game on it knowing I have 2 warranty's and let it run at 97F and this little baby just sings.

Have you tried just gaming on the machine for several hours with no futuremark and see if it still crashes? If it's not, and it's running stable even at 98F, then I feel you're letting your fears about a "potential" problem get the best of you. A problem that is covered by warranty for plenty of time to take care of it.

I'm sorry that the worry of these "What if" scenarios based on temps are getting in the way of you enjoying what is imho a Ferrari of a notebook.

I learned a long time ago to stop playing the "What if" game in life. (As in "What if my computer runs at 98F for 18 months and then dies?)

Because the "What If" game is nothing but fear getting the best of us.

I'm going to spend today gaming on this thing hard. I'll play the most taxing games I have: Metro 2033, Crysis, BCBF2, Shattered Horizon, Batman, Stalker, etc. and I'll stress the machine, leave it flat in the surface and not raised for extra cooling, run it HDMI to my 60" Pioneer and just stress test it non stop while gaming for as long as I can.

If it crashes, I'll post here or PM you. But I'm fairly confident it won't.

Conversely, if I want to ruin my enjoyment of the machine, I will run stress tests on it like Futuremark which stresses the GPU in ways games never will and really is snake oil and a hustle.

Seriously, I live in Los Angeles and have a foot in the gaming industry. And all the developers I know from that world say Futuremark and that company make a living selling software that does not offer real world data but rather stresses your system to the point of almost trying to make it crash.

Because what happens is, when you run it, you see pretty graphics and tests happening, and then your gear posts too low a score from comparable examples, or flat out crashes your gear and next thing you know you're in a panic about your purchase and are saving up or already buying new gear just to make your self feel better about a meaningless score in relation to others that you really have no verifiable proof is a real legit score anyways.

So you get caught up in the cycle and rinse and repeat. Buying new laptops, buying new video cards and Ram, upgrading your CPU - All for games designed now for a console and ported to PC to in which even middle range hardware can crush.

But videocard makers and PC makers bundle the software with their gear because they know that software is eventually going to psych you out into an upgrade, which of course means buying more of their gear.

See the hustle? And then of course Futuremark sells even more expensive versions of their software bundled with videocards and gear for the suckers that are really scared and caught up in having their ego wrapped up in their futuremark score.

It's all crap and a hustle. Especially on notebooks that all have mostly the same configuration. Futuremark makes sense if you build your desktop rig from the ground up using a variety of components and you want to see how they perform together.

But to use it as a temp monitor and gauge of whether or not you're hardware is going to fail is I believe playing right into their hustle and being manipulated by their corporate fear based products.

I bet if you loaded 10.7's onto your machine and played a game at 800/1100 and just let it run at 98F, it's not going to crash because of your temps.

And it's not going to degrade your hardware.

Continuously running a heat generating program like Futuremark will.

Now, if you've already flashed your VBios, then I don't know if it will run stable, as I feel nothing risks up your GPU or making things run unstable like that move.

If not, try just gaming on 10.7's and see what happens.

I'm sorry this issue is causing so many unneccessary concern. The only concern I'd have about that temp is my lap getting too hot if I rested the unit there. What's funny about that is, even at 97F, the Asus is way, way cooler on my lap than my Alienware M11x-R2 which tops out around 75-80F, but the way the case is designed, that little unit gets uncomfortably hot on my lap.

I myself have been in love with this unit and not having any issues with it at all other than a temp that hit 94-97.

I really love this unit. I know Asus knows that they have a hit on their hands with the G73 in the market and are well on their way to establishing a real brand. To cut corners on a product and create mass hardware failures on such a well reviewed and publicized flagship product would be corporate suicide. Will there be units that are defected in such a product line? Of course. But you have a built in 2 year warranty if yours happens to be one.

I don't believe this unit is a lemon. I believe it is a budget priced high end gaming laptop using a new case design to incorporate the latest and "hottest" (as in temp) components. As will all tech, kinks will get worked out with every generation of the product and if you have one, it should be covered by warranty.

Forums like these always attract those that have had issues with the unit as they come looking for answers to fix it. But to gauge the posters here as a reflection of overall quality vs. units sold, is not accurate.

Most purchasers who are happy with their G73's and are gaming and enjoying them don't even come here.

If the unit constantly crashes "in games" then RMA it for a replacement. If that unit crashes again "in games" then repeat the procedure until they get you what you want or you get reimbursed equitably.

If the unit is running stable, albeit high temps, but not crashing and the only way it is crashing is in Futuremark -

Well then, the problem probably isn't the G73 but the owner getting sucked into and hustled into running BS stress tests to get manipulated into buying more hardware by snake oil Futuremark.

Another thing to consider why Futuremark is a horrible gauge for hardware performance while gaming, is that in Futuremark it stresses the GPU %100 the whole time during the test. This never happens in gaming. Because at some point, you're going to finish a level or get fragged and have to respawn, etc. and it's during those moments that the GPU gets a rest and a chance to cool down before going to the max again. Just like a heart gets a rest between every beat, the GPU gets it's moments to cool down in every game.

You won't see this in Futuremark. All you'll see is your GPU stressed to the point of shutdown to psych you out and buy more hardware.

This is another reason why Futuremark is a bogus comparison tool for gaming when checking your hardware.

I trust that things will get worked out for you and others that you'll get to a place and be able to really appreciate and enjoy being able to game on this fantastic machine. A portable desktop replacement that will cost you at least $1000 more for the same thing from any decent competitor.

But if you're machine is stable at 98F, stop getting caught up in the fear machine and comparing your unit and experience to others and just have fun with it as intended and game.

It's not going to fry. If it does, you have a warranty. And if you're a hardcore gamer you're probably going to be in another machine by the end of your warranty anyways on the slight chance it dies after your warranty ends. And even then, I bet whatever it is, would be affordably fixable.

As for other GPU issues, I can't comment. But if you're machine is running games for several hours at a time at 98F and not shutting itself down, then you have nothing wrong with your unit and should enjoy.

I'm coming at this perspective as a desktop gamer, where we build rigs with things like liquid cooling because we know gaming components run hot and as so, the G73 is the first notebook I've ever owned.

Conversely, many notebook enthusiasts pay a lot of attention to heat because of comfort when using a notebook on the go.

And I often wonder, how many G73 owners that come from the notebook side of this equation are putting too much undue worry about heat; whereas my coming from the desktop liquid cooling world, expect gaming GPU's to regularly run 95F+ and not being surprised that similar technology runs that hot and stays stable. But it seems worse then it is, because the form factor of the notebook doesn't make way for something like liquid cooling, so the end user feels the enormous amounts of heat generated when the GPU is running at nominal levels.

Bottom line, if you have a gaming notebook that is as cutting edge as this, it's going to run hot.

Period.

And users should not concern themselves with it unless temps are causing system shutdowns every session after only gaming for a couple hours or less.

If that's not happening to you, then you shouldn't be on these forums and instead playing SC2 or MW2 or Dragon's Age or Mass Effect 2.

If it is, RMA it. I'm sure the wait time for return is not what it was 6 weeks ago.

Seriously, stop getting psyched by Vanatage and Futuremark and just game. If it still crashes repeatedly then, then try OC'ing it to 800/1100 as I find things run more stable for me at those clocks while temps stay the same.

If you're still having constant crashes after little use while gaming, then I'm sorry. RMA it. And instead of panicking while it's gone that the return will suck just as much, trust that it won't and that you'll get the notebook and experience intended and expected when you decided to buy it. When it comes back, it will.

I mean, if I have a trouble free, worry free, unit then there has to be more, right?

Avoid futuremark and just use your machine to game. You'll have a much more enjoyable and rewarding experience. Using futuremark is nothing but gauranteed stress and worry and headaches. Because their comparison site will always show a higher score with the same components as yours. As in they know that doing so will manipulate you emotionally into feeling insecure about your gear and push you into an upgrade all over ego.

It's nothing but a hustle that NVidia, ATI, and Futuremark are all in on together.

Don't fall for it. Have fun with your machine instead, worry free, knowing Asus provides one of the best warranty's on the market.

I trust sooner than later, everyone here will be finding the excitement and enjoyment I am with the G73. My desktop rig is way more powerful, but I love this unit so much and have so much fun with it, it's become my primary gaming rig, instead of my gaming rig when I travel.

I hope and trust this perspective may help some lurkers on the thread that are unsure or anyone else here that isn't happy and having fun with their G73.

Cheers

 

THX5334 -
Does it still work fine even if you revert back to stock speeds? 700/1000

 

Not in my experience.

But that's only happened once and has not been repeated.

It happened a few weeks ago when playing BFBC2 on 10.6.

It was the first time I saw my G73 hit 97F. I started to trip, I windowed out of the game, reset my MSI Afterburner to stock clocks, went back into the game and within a few minutes I experienced the first crash I'd had in weeks. The last not happening since I was experimenting with the clocks on my GPU and crashes are expected.

I went back to 800/1100 except the one time to try 705/1100 and found I was getting the exact same temps as 800/1100 but losing the performance of those clocks.

I honestly haven't used the stock clocks much at all, because for me, the performance gained in overclocking the GPU from stock to 800/1100 was substantially better. An increase in 5-15 fps in games across the board. And for that one time, also seems more stable.

My stock clocks still gets the unit hot into the 90's and up to 94-95. The overclock adds a couple of degrees of heat, but it runs stable and the fps gain is considerable enough to be worth those extra degrees for me. That overclock is the difference between playing a game at Normal vs. High graphic settings, or ever more, High vs. Very High.

I can try going back to the stock clocks, but I also suggest others try OC'. Even to 705/1100. Maybe the stock clock speeds just aren't getting enough juice at those levels to stay stable? Maybe even that +5 increase on the 700 core clock is the difference between everyone's crashing and not?

I run my rig at 800/1100 on 10.7 using MSI Afterburner on bios 209 with the stock VBios. I leave the CPU turbo setting to Turbo as I find that actually gives me gains as well.

I play mainly FPS and RPG's. I don't play RTS at all(though am considering SC2)

I have 60 games in my steam library, if they're not one of those genres, then it's probably something you can also get on a console or I also play one MMO - Star Trek Online.

All of these games run fine and stable for hours at a time.

Everything on my machine runs Platinum, the only caveat being that I'm definitely in the 95 club with a peak of 97-98F.

But when it does, I remind myself it's a "gaming" notebook and keep on gaming. Having decided I'm not going to hook into other's fears about the machine and go about my way. And if by chance my own criteria for having a faulty GPU occurs - constant system shutdowns during gaming while under the same settings as always - then I'll hit up Fry's for my warranty or Asus for an RMA.

I'll try the stock clocks, but would also encourage others to try 800/1100. If your machine stays stable with it, the performance gains are worth it.

Cheers

 

THX5334 -
I will try your clocks because when I run normal clocks with the 10.7 ATI Package my games will either freeze or it will gsod for 3 seconds the it will get a checkers board look to it.

 

When mine last crashed during BCBF2 it was during a level set in a snow/winter environment - single player campaign.

The screen went to all white, but you could still hear the game running through the speakers, so it was definitely the GPU.

Everyone should try the clocks. If they make your machine stable and add performance gains then it's a double win.

If that fixes your issues but you still have high temps and are worried, and expect temps one would get from a $500 Lenovo, instead of the normal high temps desktop gamers are used to that this GPU is definitively modeled after, then stop. Remember that from that perspective it's operating within spec. and you have a warranty that covers 2 years. And how often do PC gamers use hardware longer than that?

If the high temps still have you stressed out, then your problem isn't your notebook. Your problem is that you're addicted to fear and worry. Quite common as mental health scientists have proven that those emotions create a stimulation or "arousal" in the very same brain areas and ways as anything sexual that can trigger a rush that creates a dependency to lean towards those feelings. Nowhere else, do you see this fear mechanism play out than with PC/Notebook/gamer/tech hobbyists. As the purchases usually require a significant amount of income that could go to more essential things combined with a regenerating hardware cycle that creates a want/anxiety for the new. It's a common and sucky thing we enthusiasts go through and why we often find ourselves spending more time here than on the tech we're discussing and using it as intended. That's why I often buy the extended warranty even though a part of me know it's a hustle. It buys me piece of mind from things like many are going through here. I like to come here because there's also enthusiasts and those with much more knowledge than I that offer advice and tweaks or news that make the system run even better.

Gawd, I've been typing these posts out on my A1. I'm a writer, and I just love the feel and am addicted to the feel and click of this keyboard. As I'm sure length of these posts can attest. Man I lurv this rig. I really do. I've wanted a notebook for so long but had to have one that could game. And I could never find a happy medium between the performance my graphics-floozy side demands combined with a price I could afford combined with a size that albeit large, doesn't feel like the bulky Buick of a Clevo or Malibal which kept me away from portable gaming rigs. But the G73JH actually looks like a notebook with the performance to match both of those brands and Alienware at approx. $1000 less than any of those brands. And that's for the premium A1 config with Blu-Ray, not counting A2, X1, etc.

The Asus G73JH is a Pimp-Jedi gaming machine. Believe it.

 

whats the safest temps for the memIO? mine reaches 104 accordung to AMD GPU Clock Tool after an hour of playing GTA 4 EFLC while the core and shader are at 90 and 94

 

If you can game at those temps another hour or two with no shutdowns, why should you care? You have a warranty right?

 

THX5334
First, its not Futuremark, but Furmark.
Futuremark is a company that makes the so well known 3Dmark software that doesnt even come close to stressing the gpu like Furmark does. I think thats the first thing you mixed up. But unlike you i value their products as well, not for bragging purposes, but simply to gage if a machine performs as well as it should (there are smaller deviations, but for purposes to compare two identical systems or systems with same gpus to see if both perform similarly and as they should its great).
And Furmark is a completely safe software to test your GPU - all it does is make the gpu run at 100% load which is pretty similar to what some demanding games and software do to it (crysis is a good example of such a game - it pushes the gpu to 100% and levels can be long and the GPU will heat to pretty similar temperatures and wont cool much during level loading - might drop a few degrees for a minute but will go up fast again when the new level starts). If your GPU has a problem however - either stability of overheating it will show up fast in furmark (it will GSOD or overheat).

I posted several times that we who are testing this (myself included) have GSODs on Furmak, as well as in games and other software (but Furmark makes it show faster).

If the system doesnt work stock im not satisfied with it. Even if i get ALMOST 100% stability in games by overclocking, or flashing a custom undervolted/overvolted/underclocked/overclocked vBios i find that unacceptable (perhaps, but only perhaps i could swallow it if those solutions worked to get us 100% stability, but the dont, they just improve the situation a bit). A system this expensive has to work without a problem. Id also make a fuss with a less expensive one, but those that i have actually work just fine unlike this brick.
Its not about synthetic scores, we never discussed that, but simple stability issues. The thing DOESNT WORK.

No, it wont cost 1000$ more, go configure the w860cu the same and get your facts straight. It costs more, but not so much more and it offers a lot more for that price (stability, build quality, better cooling, esata, firewire...). And even if it did cost 1 grand more, at least you get a working system, here you paid a ridiculous ammount for a laptop that doesnt even work and you cant use.

We were talking in degrees celsious, not fahrenheit. 98F is ridiculously low (36C), the gpu IDLES at around 60C. Some users have a gpu that runs over 100C at full load (212F) (i also had the first stock pasted gpu like that) or at least over 90C which is certainly not good long term and will cause it to slowly degrade.

If it fried and i had to wait 2 months for it to be repaired like some users, id loose it. 2 weeks is overkill and Asus is known for how painfully long they take with RMAs. Its never nice for something to die, things should be made to last. And what if it happens just over 2 years when its not covered by the warranty anymore? Notebook components are by no means cheap. If youre fine with wasting money, go ahead, but i dont intend to pay for something i shouldnt have to. Some people expect to keep their laptops for a long time and this laptop certainly isnt promising in that regard.

There are users who get GSODs after minutes, so they cant even start to play, some get them during longer gaming sessisions. So it is a problem and thats why we are here.

I did get a gpu replacement already and its the same, and all the other notebooks the techs tested in the warehouses (Ken and a local tech) also had the exact same problems so i dont feel like taking it apart in vain again.

You may be lucky enough to be relatively stable on your overclock and havent stressed it enough yet, but i bet you if you run games long enough youll gsod - but youll see it faster with furmark if you really want to know just how (un)stable your system is.

Dont make me laugh talking about the warranty. The Asus warranty sounds nice - 2 years + 1 year accidental, but what good is it if they wont even fix such a huge problem?

800/1100 overclock also wont get you a 5-15 fps increase in all games. Ive tried and with crysis gains are practically squat with that.

You may like your machine and youre entitled to your oppinion, but its a fact that this laptop is cheaper for a reason - it is the worst of the competition when it comes to quality, performance and stability.

It seams you didnt even read our posts thoroughly and dont know what youre talking about. Youve just majorly embarassed yourself.

 

Not to rain on your parade or anything... but...

The G73 is a great machine. Great hardware for the price, very silent, the dual-fan clamshell fans do a great job at heat dissipation and noise control. Then there's the thermal protection on their Asus manufactured GPU card. I mean, what a place to fuxx-up. My Dragon Age temps on TSS1 (MemIO) would hit the 96C mark. Then I went ahead and ripped open the unit and redid the thermal paste myself. The same game, at same clock speeds, and same settings now hit 73C. That's -23C folks! Yes that's right, which is about what competitor units (Clevo) are getting with the HD5870M.

If you want more info, hit the Asus Gamers Notebook forum and look at Chastity's Repaste Experience for pics and numbers.

Apparently my GPU was assembled by the two techs in my sig pic.

 

Just one thing is bothering me Chastity - noone else has been able to get such great temperatures and some pretty savy people have redone their TIM and more on their G73JH (including myslef, and i consider myself pretty expirienced by now taking apart more than different notebooks for myself and friends to reapply TIM, clean them or upgrade components or even do something like successfully cram an unsuported gpu into a laptop).
I took the card out and fitted the heatsink on to see how much TIM i needed (i used the same TIM - MX-3 on the 5870 in D900F and there temps were a LOT lower) and if contact and tension were good and since my TS1 is within 2-3C of TS0 i think i did a good job, but my max temperature is 94C on furmark. Our ambient is very different too though - your 20-21C vs my 27-28C.
Its great you got such good temps, but perhaps dont expect others to get it too, since you seam to be a pretty rare case. The GPUs do differ themselves a bit in how hot they run as well, you should know that too.

 

I know this, but I would like to say that Mr. Lee also gets some comparable temps with his repastes. (I believe he has some Furmarks hitting 82-84C) SO I have to believe that people should be able to hit the same temp region.

I know you know this, but I'll elaborate for the other viewers:

Thermal paste is meant to just fill in gaps between the heatsink metal and the actual processor, CPU or GPU regardless. The heatsink should be brought in as close as physically possible, without crushing , so the amount of paste should be used sparingly. The majority of the heat is usually centrally located on the processor, this is why paste is usually applied in the middle, and allowed to spread out. Don't worry about the edges of the chip: getting the core area covered is the main goal. So having the gap just barely filled with thermal paste is optimal. The paste will be thin to help transfer heat to the heatsink. If you have more than is needed, then the paste can act as a thermal insulator, and less heat will be transferred over a period of time. This is the condition I believe is happening with our G73 GPUs: too much paste and gap between heatsink and gpu.

With my own pasting, I guess I got a really good paste thickness, and an optimal heatsink gap for really good thermal transfer. I have noticed the air coming out the GPU port is warmer than before. (which means the thermal protection is doing it's job. More heat coming out than is being kept on the GPU)

 

How long of a furmark run are you talking about before you see a gsod? days? weeks? I've yet to gsod on mine and I was running mw2, metro 2033, bc2 for a few hours at a time. I didn't try furmark for long but I'll try a longer run to see if it throws a gsod.

 

With a 6-7 degree lower ambient, you get a significantly cooler system. It isn't simply a 6-7 degree drop, it's the fact that the two systems has 6-7 degrees of difference between them and the air surrounding them.

So a 6-7 degree difference (lower) in ambient is going to allow for much more heat to be redirected to the air in a smaller amount of time... leaving 10-15 degree differences in GPU temp.

 

Umm. i applied AS5 on my GPU and I gota higher temperature, 91C furmark.
Differences AS5 vs IC7 2c to 5c max
AS5 cure time of 200 hours. 2c to 4c least some people.
20c to 24c ambient temperature.
91 - 5c(ic7) -4c(200H)= 82c as best scenary

Chastity and Mr. Lee use some type of cooler to help reduce temperatures.

Or the thermal pad are not making good contact and do not pass heat to the heatsink.

I think that because when I underclocked the memories in the hd4890, the temperature was reduced over the whole GPU.

 

Used a cooler? That's not a very nice thing to say!

Actually, the only thing I did was have it sit atop my LapChill, tho not plugged in. (I didn't feel like running 4hr + tests resting on my bed.)

As for your temps, AS5 takes a LONG time to cure. You should see improvements as the paste spreads out and gets thinner. Also there are other variables: how flat the bugbears left your heatsink, how much gap you attained when you tightened down, and how much paste you used. The pads shouldn't affect much, all the sensors are on the GPU.

Also, did you give the heatsink a twist before locking it down? That helps remove air pockets in the paste. Also, did you clean the surface really good? Residue from old paste can be an insulator.

 

Sorry it was more than a question, than an affirmation

Cleaned de gpu and heatsink whit a isopropyl alcohol.