mauasatto

First of all let me say this is an awesome site!

In my desperate need of help I have searched a lot of forums and have finally decided to post on this one.

I have recently bought an Asus v9999 GE, and the problems began...
Doom3 ran just fine.
However, with other games (NWN), as I tried to maximize the graphical settings (e.g. add antialiasing, etc), the GPU and RAM would sometimes overheat, with the Fan never passing the ~3,600 RPM when set to "manual - 100% power".
As I am very pedantic, I decided the overheating occured because the fan was malfunctioning (after all, it has a graph of upto 10,000 RPM and it just uses 35% of it...) so I thought this was the problem, and I sent the card to Asus for a replacement.

Now I got a card back. I tried running "Beyond Good and Evil" yesterday for the first time (with full settings on 1024/768), and I have gotten some extreme overheating just a few minutes into the game: GPU 89c and RAM 79c.
Today, the game runs and freezes the computer after only 1 minute of running! - I get a black screen and the computer freezes!

The funniest thing is how UN-helpful Asus are: I have contacted them several times just to get silly questions like "are you over-clocking?" or "have you installed the latest drivers?"...

I am not and have never over-clocked. I use SmartDoctor for monitoring only.
I have the latest drivers and smartdoctor installed.
My system details:
MB: Asus P4P800-Deluxe
CPU: Intel Pentium4 3ghz (800mhz)
CPU-Fan: Intel in-a-box tower+fan running at ~4400 RPM
Memory: Micron DDR400 1GB
Power supply: 420 W (unheard-of manufacturer)
Case: Ci-6A19 + Extra 12" fan at ~1300 RPM
Hdd: Maxtor SATA 80 GB.
OS: Windows XP pro.

Please help me.
What can I do to stop the overheating, the system freezes and black screens?

The only possible important detail I can think of is my power supply, which is 420W but from a cheap make. I *could* get a Thermaltech 420W which would be much more stable and reliable... but would this solve the problem or would I just be spending $70 (this is what it costs here in Israel!) just to see the overheating and freezes repeat themselves??

...oh, yeah - I forgot to mention one last thing: It has just become winter here in Israel, and ironically the overheating was less problematic during the summer than now...

Praetor

Define overheat

Thats not overheating. The 6800s can breach 100C without blinking (or throttling, unless youve set them to)

Annoying yes but you have no idea how neccesary it is to ask those questions.

Im not convinced its overheating -- ive ran a 6800GT and 6800U without a fan along with Particule Fury (albeit not for very long) ... now THAT's overheating

Its good that you mention that. How stable ARE your rails? (use ASUS probe to check ... have it start loggin and then play some games until it crashes)

__________________
ASUS P5K Premium WiFi-AP, Q6600@3.7 / ASUS P5ND, E6400@3.8
4GB OCz Platinum XTC 8500 / 4GB CorsairXMS2 6400
5x500GB Seagate 7200.10 / 2x500 Seagate 7200.10
OCz 8800GTX 768MB @ 630/800 / 2x Galaxy 8800GT SLI

mauasatto

Wow! what a quick reply! You caught me off-guard there...

I get the clue... maybe 80c on the GPU is not exactly overheating.

I gather you are referring to the GPU. What about the RAM? Can it also breach the 100c?

Ohh.. I wouldn't try that..
Nevertheless, I should mention that yesterday I got the overheating message from SmartDoctor only after I got out of the game myself, after having played a few minutes. And today it gives me the black-freeze during the demo, not even getting to the game itself, not to mention that I had to turn the overheating alarm "off while playing 3d-games" in smartdoctor just to prevent the smartdoctor from interfering with the game process.

Are you sure my case is not overheating...?

Then what could it be?

I should mention the following:
Yesterday I was running on the Asus 62.11 drivers with SmartDoctor 4.54.
But today I am running with the Asus 66.72 drivers (in this 66.72 version Asus reverted to the Nvidia 6800 drivers, so I also had to install the Asus "Enhanced" Driver and I did) with SmartDoctor 4.57.

Could *this* have anything to do with it?

Wow... I don't know... I don't even know what "rails" are...
I will search this "Asus probe" you mention. I gather by what you say that this is some type of logging program. I believe I will manage it. I will post the results in a few minutes.

Meanwhile, thank you so very much, I hope your help will get me through this!

mauasatto

Praetor:

Okay, I've installed the latest Asus probe and made it log the history, polling every 5 seconds.

I tried to get the computer to freeze, but the Asus Probe program would not allow this to happen: I got a "floating point divide by zero" error and a following "exception EaccessViolation" error - all errors from the Asus Probe which turned to focus over to the Asus Probe program, thus minimizing the game and (so I believe) preventing overheat, because its enought to minimize the game for 5 seconds and the temperatures drop more than 15c.

So I couldn't manage to get a freeze now, because each time I re-ran the Asus Probe and returned to the game, not a minute passed without some new error from the Asus probe.

The Power polls were did indeed manifest some fluctuations, but I believe a certain amount of fluctuation is normal, is it not?
Nevertheless, here is the most different, highest poll of all, taken quite frankly *before* running the 3d games:
+12v: 12.403
+5v: 5.187
+3.3v: 3.392
VCore: 1.616

Eagerly expecting your reply to my last 2 posts,
Mauasatto.

Praetor

You have the ASUS V9999GE which uses GDDR3 ... the nice thing about GDDR3 is that temperature isnt so much an issue (not that it ever was, but even in comparison with DDR). If the core isnt hitting 100C then obviously the memory wont be even hitting 80C

Screw smartdoctor (i never liked them stupid proggies anyways heehee). What does the nVidia temperature sensor tell you?

Again screw smartdoctor.

Ive not stated that but for the record, what ARE your temps?

While we're on it, screw the ASUS drivers Get the real nvidia ones not silly ASUS ones (which are just ASUSized versions of the nVidia drivers). I had a very good experience with the 61.77s and now im using the 70.41s ... the current 66.93s are good too. You can get them from http://www.nvidia.com/object/winxp_2k_66.93

Rails = the voltages lines on your motherboard, 12V, 5V and 3.3V ... if your videocard isnt getting good power then its not gonna be happy.

This is more indicitive of a system/memory problem rather than speciifcally the videocard/probe problem.

Screw games. Run particle fury for a bit and ramp it to 120K particles (and while you're at it, whats your fps at default 32K and 120K just for reference sake). You can have ASUS probe running (because particle fury isnt a fullscreen program) and have the nVidia temperature thingy running so you can check the tmps.

That seems stable however its not entirely usefule because, as noted, its before you started playing. You might consider trying to run MBM5 (motherboard monitor5).

MBM5: http://www.dvhardware.net/modules.ph...page&pid=3%20?
Particle Fury: http://www.setiathomescreensaverspee...l/particle.htm or http://www.majorgeeks.com/download158.html

__________________
ASUS P5K Premium WiFi-AP, Q6600@3.7 / ASUS P5ND, E6400@3.8
4GB OCz Platinum XTC 8500 / 4GB CorsairXMS2 6400
5x500GB Seagate 7200.10 / 2x500 Seagate 7200.10
OCz 8800GTX 768MB @ 630/800 / 2x Galaxy 8800GT SLI

mauasatto

Praetor:
Today is Friday here Israel, which is kind of like your Saturday.
I have to get going to my Friday night dinner @ friends house pretty soon, so before I download Particle Fury to test those FPS rates, I will reply the message first:

Actually I wouldn't count on it... the highest temperatures I have seen were:
GPU: 89 celcius
RAM: 79 celcius
*BUT* these were the temperatures I saw after EXITING the game. This means that during the game the temperatures must've been higher.

Another bothersome issue is the idlying temperature of the GPU...
As I've mentioned, during the summer the temperatures were lower, for some obscure reason...
During the summer, the GPU was idle at ~54 celcius, and the RAM at ~44 celcius.
Then, as I've told you, I sent the card to Asus and got a new one instead, for I was not happy with the Fan's RPM (never above 3667 RPM).

Now, in the winter's beginning, the GPU is idle at ~66 celcius and the RAM at ~54. This is baffling... could I have gotten as a replacement a card that works worse than the previous one? Because, by the way, the Fan's RPM *continues* to be always lower than or equal to 3667...
But all these polls come from your favorite pal, SmartDoctor, which I have begun to hate myself... I mean: a process of ~14MB just for polling temperatures and crap like that?! Come on...

Hmm... I have never tried any other monitoring program on this card. I know that Asus have tampered a bit with the whole configuration around the 6800 on their v9999 GE, so I was hesitant to use any software but the intended ones by Asus... but nevertheless I will try this nVidia program you speak of, right before I try that particle fury.


I have mentioned them above, as I have mentioned my concern that the values I saw were un-reflective of reality, for they were seen after exiting the game.
What do you make of these values (GPU 89c, RAM 79c)?

I wonder: if the RAM and the GPU can sustain over 100c, exactly how much?
and I wonder: if it is so, then why did Asus program the SmartDoctor to alert by default at values of 80 for GPU and 70 for RAM?

The most important question is probably: do you think it is safe for me to go on playing a game (assuming it does not hang my machine) all the while ignoring such values?



I have never tried using the original nvidia ones, for the same reason that I have not tried the original nvidia temperature sensor program. But I will do as you say and attempt a whole removal of the drivers and SmartDoctor for the sake of the original Nvidia ones, even if it means ignoring my concerns.

Probably a software issue. I highly doubt that there is anything wrong with my memory. I use my machine on a different boot-setup for music, and I completely abuse the memory with gigabytes of samples and wavs. The secondary boot has never manifested any problems whatsoever, and is configured at a much lighter, good-for-work-and-not-for-play state - which leads me to believe it is likely a software problem. Possibly something to do with this dumbass norton antivirus which I stubbornly refuse to remove (in spite of being behind a physical firewall machine and a router... I guess all these things I'm saying, little by little - the fear of using nvidia programs over asus ones, the stubbornness by leaving NAV installed - they all kinda' lead to me being a chicken... but the fact is that the secondary boot has my music, and I am reluctant to endanger it in any possible way...)

Actually the program managed to do some polling in-game. The reason I specified the before-game values was that they were more abnormal than the in-game ones. Actually, while in-game, the polling results were much more stable and closer to the ideal voltage values. That is why I did not mention them.

This sounds ideal.
This is what I will do as soon as I get back home from that dinner.

Regardless, I would very much appreciate your opinion of the temperatures I have given you and the questions that followed.

btw: how's the weather in Canada now? I bet you could overclock your whole machine so long as you turned the heating off...

mauasatto

I've searched the nvidia site - and I have not found any nvidia temperature sensor. Only something for nForce motherboards. I've also searched google - no results for "nvidia temperature sensor".

And the nvidia drivers version 66.93 do not support Gforce6800 GE. They support the GT and plain 6800s, but no GE.


What should I do?

Also: do you think it is safe to go on playing while ignoring the "GPU=89c" and "RAM=79c" warnings?

mauasatto

Praetor:

I have not found any Nvidia temperature monitor programs.

I have tried particle fury. At 120,000 particles the card's temperature did NOT go over 68 celcius on SmartDoctor. I was dissapointed... anyways, here's
what you wanted to know:

Particle fury results:
120,000 particles FPS 25.7
I understand frames per second, but the rest of the details are incomprehensible to me:
3082 kp/sec [average: 2677 kp/sec] [peak: 3135 kp/sec]

at 32kparticles the FPS was something like 97, don't remember exactly.

I also would really like your answer:
do you think it is safe to go on playing while the GPU is over 90c and RAM over 80c?
I have nowhere else to turn to, please help me.

Praetor

A tad higher than I might have expected but nothing to be concerned with at all.

Idle temp for stock cooling here is roughly that temperature -- and quit worrying about the ram temps. They're fine.

Why is the RPM an issue? Does it say somewhere that the rpm has/should be a certain value?

.
Screw the software and drivers that come with the ASUS card. Use the nvidia drivers (which are newer). You can get them from here http://www.nvidia.com/object/winxp_2k_66.93

Nothing at all to be concerned about ... unless your ambient temp is like -10C or something... what is your ambient temp (room) and ambient temp (case)?

1. Quit worrying about the ram
2. The 6800 series automatically starts to throttle itself when it hits 120C ... which leaves you plenty of room

Because it's stupid.

Set the threshold higher.

And IDS and router wont protect you from the inside

Get over it (seriously). I had a V9999Ultra Deluxe a couple weeks ago ... didnt bother installing the drivers from ASUS... went straight to nVidia. Think about it this way: you look at eBay or whatnot ... they sell OEM cards (cards without CDs and boxes and fancy drivers) ... what do you think the people who buy those cards do? The likeliness of each one of them having a copy of ASUS's driver CD is remote and granted they could goto the ASUS website (and navigate through the mess IMO) but most just goto nVidia.com and grab the latest.

That's good stubborness.

The heating is off. The windows are wide open. The ambient room temp is -6 to -12. Case temp is in the 10C-15C range.

LOL. Well first you have to overcome your stubborness and install the nVidia drivers. Heehee. Then you can get to the temps tab (Display --> Settings-->Advanced-->GeForce 6800 -->Temperature Settings)

They support it (otherwise I wouldnt be reccomending you look now would i? ). The GE stands for Gamer Edition -- an ASUS name. Not an nVidia name. Chip in there is a GeForce6800. You'll be fine.

As long as "over 90C" means "over 90C and less than 125C"

__________________
ASUS P5K Premium WiFi-AP, Q6600@3.7 / ASUS P5ND, E6400@3.8
4GB OCz Platinum XTC 8500 / 4GB CorsairXMS2 6400
5x500GB Seagate 7200.10 / 2x500 Seagate 7200.10
OCz 8800GTX 768MB @ 630/800 / 2x Galaxy 8800GT SLI

mauasatto

Praetor:

First let me say thank you very much for being so patient with me..

I've reinstalled the nvidia drivers and looked again: I've found no such option as "temperature settings" where you said it would be. I've checked every other possible place and have found nothing.


Well, actually - yes. Asus have said to me in one of their mails that the RPM of the 6800 GE should be around 4,500.

I've gotten them the last time you suggested them.
They are installed right now. Only thing is: I can't monitor the temperature and this is unsettling.

I would say roughly:
Room: 15c (right now)
Case: I don't know what you mean by "case", but the CPU is 34c and the MB is 30c. The metal casing outside feels at about 10c. Does this help??

This is a reassuring fact that I did not know.
It does not however solve the possibility of such heat inside the computer affecting other things (my 600$ audio card, for instance, or any other plastic cables...)

Well, the StupidDoctor has a maximum threshold of 90c for the GPU.
I've just tried running the game with this threshold, and I passed it in three minutes.
Nevertheless, the threshold is not a problem: I can turn the alarm "off while in 3D games", as long as I feel it's safe to do so.

You see, Praetor, this whole mess is being created because I am trying to use the same computer I use for my music - to play games. If this was not the machine I worked on, I would just try and see what happens! Everything is under warranty - except for my Audio card and my work... which are the most important things...

You're right.

Jeez! How can you sit in that room? Wearing a bear-fur coat?

The following options appear in the tab you specified:
-Screen Adjustment
-Display Mode Timing
-Performance & Quality Settings
-Color Correction
-Video Overlay Settings
-Troubleshooting
-NVRotate
-Refresh Rate Overrides
-Screen Resolutions & Refresh Rates
+Desktop Management
-Profiles
-Menu Editing

"Temperature Settings" is not one of them, nor did I see it under any of the above options...