I've had this horrible error hounding me for literaly three years now, people call it the infinite loop error. For some it happens all the time, for me only in certain games (LFS included unfortuantely)

It seems to be related to my graphics card, the error is located in the graphics drivers, but the wierd thing is it seems to have followed me across complete hardware changes.

Over the past two years I've done everything I can think of. I've changed my GFX card (made it WORSE), ram, and PSU, I've even gone as far as to change my motherboard, in hopes this will go away.

The only thing I haven't done is changed my CPU or case!

The symptoms are pretty simple, the whole computer hard locks, nothing but the power or reset switch will work, and the screen goes black and sleeps.

It started on my Radeon 9000, but it was never a real problem, now with the HD 2600 pro AGP I was just barely able to afford at the time (I'm a poor bastard), it happens all the time. It's not limmited to ATI cards though, as people experiance it on Nvidia cards as well.

I think its heat related because no matter what my case temp always hovers around 40ºC at idle, And before I did a fan speed tweak on my GFX card it was always at 55-58ºC idle. now I've got it idle at 51ºC, but I still think thats too high.

I'm thinking of getting a bigger case, because I've been told my case is too small for all the hardware I've got crammed in it. I've got two 80mm fans on the back, an 80mm fan on the side, and the biggest CPU cooler that will fit in the case and it still idles at 40ºC, and the CPU still gets very hot.

Do you guys think a bigger case is my only option here? I mean the number of fans I'm running is pretty rediculous and it still runs hot.
Alternatively, do you think that could be the problem at all?

Aw man, sorry to hear that, I had the same issue before on an old computer. If it is running hot, and the HD light is on, yeah, most likely it's now a paperweight. It can happen for many reasons, most likely a process was running while the computer lost power, generated errors, and couldn't be terminated, or has items that cannot be solved. It's running hot, because that process is unfortunately consuming all your ram / cpu resources. =(

I just salvaged what I could when it happened to me, you should do the same before it does damage to your new components. Keep the cd/dvd drives, graphics chip, ram cards, if you have a wifi card get that too, but pretty much the hard drive is toast, and no matter what you replace, that HD and possibly your motherboard have already been permanently affected. Overheating alone does major damage to the circuitry.

Sorry, I know it hurts, but better to cut your losses now, and buy a new computer, just save the extra components if you need them later. I wish ya the best of luck bro. =)

i had a similar problem years ago. my pc would lock up at any time when doing anything. black screen, nothing worked. turned out the cpu was over heating and not shutting down to protect itself.

I replaced the Motherboard and formated the HDD, so theres no chance that there is a thread stuck in the CPU causing the overheating.

Its ran hot since the day I got it.

Quote from dadge :

i had a similar problem years ago. my pc would lock up at any time when doing anything. black screen, nothing worked. turned out the cpu was over heating and not shutting down to protect itself.

That sounds more logical than Pablo's answer.

If thats the case, than I should probably get a bigger/better Case then. since I'm already running a very good heatsink/fan on the CPU.

Did you ever see an error, like XXXX.dll got stuck in an infinite loop?

IMO ignore Pablo's post.

If it performs well some of the time then the graphics card is most certainly not a paperweight. If it had been damaged by excess heat it would show those signs constantly. 60 degrees is not a very high temperature for a modern graphics card. 40 degrees is not a high idle temperature for a modern CPU, especially if you have a smaller/less effective heatsink. If you read the hardware specs you'll probably find the maximum recommended temperature for your graphics card and CPU are 75+ degrees C.

Your problem sounds a lot more like a driver or an OS issue than a problem with hardware. Are you still using the same type of OS you've been using for the last 3 years (e.g. sticking with Windows XP) or have you changed?

Can you tell us your full system specification please? CPU, PSU, Motherboards, monitors etc including the different ones you've tried.

I can give a full list of hardware, but I don't remember details on some things

starting system:
-450w PSU (crappy brandless garbage)
-ECS PM800-M2 motherboard (via chipset)
-512mb Kingston DDR400 ram
-Radeon 9000 All in wonder AGP 4X
-Pentium 4 Prescott @ 3.2gh with HT enabled

Upgrades in order:

-500w PSU Brand is hidden by case unfortunately, but I made sure it was good when I bought it. I bought it within a week of getting the computer.

-zulman 7000 CPU cooler. best one at the time.

-second Kingston 512mb stick, old one failed shortly after.

-switched from windows 98 to XP SP2.

-very big gap in hardware changes, lots of driver changes.

-Radeon HD 2600 512mb pro AGP 8X.

-2gb matching pair of Kingston 1gb DDR400.

-MSI MS-6788 (intel chipset)

It was a whole year before I noticed a problem, it had crashed, but a passed it off as windows being glitchy.

Durring this entire time I have switched to every concivable driver version and it even carried over to XP from windows 98.

Also, my CPU has hit +80ºC before, and I can't monitor it while gaming so it might be doing it still.

I heard that VIA has problems with ATI AGP cards, and just AGP in general, so I changed to an intel chipset.

Ok, thanks for the details.

Are any of the components overclocked at the moment or have they been overclocked at any time in their lifetime?

Quote from DragonCommando :

It was a whole year before I noticed a problem, it had crashed, but a passed it off as windows being glitchy.

You didn't notice problems before swapping to the 500W branded PSU? If not why did you switch the PSU? For peace of mind because of the no name one that you had originally?

What I would try:

Stress testing the RAM with Memtest86. Download the bootable ISO, burn to CD (make sure you create a bootable CD), boot into it and let it run the diagnostic for a few hours. Most people would say 6+ hours...leaving it on overnight is a good idea if you can. You've already said one stick of RAM was defective, how did you determine that? Was it through a diagnostic utility like Memtest or was it by swapping the RAM and not having problems?

Stress testing the CPU with Prime95. Similar to Memtest but most people recommend an ever longer test period. Do 24 hours if you can...you run Prime95 from Windows so you can still use the PC at the same time as testing.

I switched PSUs because the no-name one wasn't all that future proof, they took it as an exchange and I only had to pay the difference. I never like buying no-name brand parts, it always leaves too much to be questioned.

I have never even tried overclocking anything, this computer has always run too hot for my liking. I would be too afraid to do damage.

I have memory tested every single memory module I have ever installed in this computer. I left it for about 12 hours every time. It only failed once, and that was for the original module the computer came with.

I will try prime95, I never stress tested my CPU. I'm pretty sure I know what the log will look like when its done though. Right now it's idling at about 56 degrees, And thats with the case open.

Edit 1: The CPU is already at 70ºC and its been running for only 5 minuts.

Edit 2: I tried to run the CS:S video stress test in window mode, to see what would happen to my graphics card temp. Ran nice and smoothe, untill it froze. something tells me it has to do with my CPU being at 70ºC because I ran the test on an idle processor and it never crashed.

Sorry, I thought the issue he said was an infinite loop, lol mai bad. =X I had a Compaq that locked up on me due to my sister excessively ddosing with it, it was such a waste of an otherwise running machine.

Quote from Pablo Donoso :

I had a Compaq that locked up on me due to my sister excessively ddosing with it

If you can, borrow some branded powerful PSU to eliminate the possiblity of weak PSU (I've had this problem with PC similar to yours). Maybe updating your MB's BIOS is also worth a try...

I actualy already updated the bios, and the PSU was changed for a good branded one the same week I got the computer.

That`s a whole lotta heat you got there.
Have you tried cleaning the CPU with cotton swabs & isopropyl alcohol, then reseat the heatsink? with a new layer of thermal compound, something like as5, mx-2.
Also make sure the case inside it`s dust free. 40c ambient temp seems too much for a case with some airflow.

With "two 80mm fans on the back, an 80mm fan on the side" you can try setting up some negative pressure, if you have an "el cheapo" case, you can ghetto mod it (read: cut it) with a top exhaust fan, larger than 120mm if you can (hot air goes up). Goes without saying you keep the case closed, if you want any airflow.

And what do you mean by saying it`s related to graphics drivers, does nv4_disp / ati2_dvag dlls show up on BSODing?

Try this, skip to the "resolution" chapter, maybe you can find out more about what`s causing it with dr. watson.

Need to know more, what does the BSOD actually "say"? Or it just freezes, no bsod?
The total lock-up can be caused by overheating.

Have you tried re-installing windows? It wasn't clear from your post.

@Jakg

I have reinstalled windows several times, even different versions.

@all

I have completely cleaned the case, replaced the thermal compound, tried more powerful fans, the works. Apearantly prescotts are well known for producing alot of heat, 103w of heat dissipation. And it doesn't help that my case is a pretty cheap one.

I have semi-ghetto modded it already, I cut a hole in the side, because it didn't come with a side mount fan from the store. I put an 80mm industrial helical prop fan in the side. It doen't have a speed sensor, atleast not one that is compatible with a PC motherboard, but it causes the CPU cooler fan to rotate at 3000rpm instead of 2600rpm. It has a through put of approximately 55CFM. My two rear vent fans are 32CFM each as well.

I realy think I'm wasting my time with this case.


It doesn't BSOD for some reason, it used to when I had Windows 98 and very early on with XP.

The error is "The device driver ati_(insert name here).dll for device video 0 got stuck in an infinite loop". The actual DLL name changes from one driver version to the next. I get this message in event viewer, but only if I leave the computer locked untill it auto reboots, or the sound stops looping.

I realy hope that the final fix is getting a better case, I realy never thaught it would cause this much trouble, but after stress testing with prime95 I realize I have a realy big overheating issue, and I can't afford to replace much more than the case.

I was offered an xblade case yesterday (right after posting even) for $65 without fans or PSU, I have all the fans I could dream of since I collect all kinds of junk, and I am more than sure my PSU is still good since it actualy runs closer to specifications than I have ever seen in a PSU.

I just don't want to buy the case and find out heat wasn't the issue, although the case would still make a realy nice one for my next computer.

Quote from DragonCommando :

I just don't want to buy the case and find out heat wasn't the issue, although the case would still make a realy nice one for my next computer.

Put all of it on a table and monitor the temps. again. This way you´ll know if you need another case. I doubt it.

I've tried every driver version and variation compatable with my card. I even added my card's ID lines to the first version drivers, which gave the same resaults.

It appeard for a while that different drivers would crash at different times or in some applications that others would not. But what I have found is that seems to be completely related to my habbits and not some differences in the drivers. I get crashes after longer periods of time if the computer is freshly turned on.

If I play LFS it still crashes, which leads me to belive its not the graphics card or drivers, LFS uses such simple DX code it can't be causing problems when I can run 3d studio max perfectly fine indefinately in DX 8 or 9 mode.

Plus it seems to be completely API independant, most 3d driver issues are API dependant. I play a port of Duke3d called Eduke32 that runs OpenGL and it still crashes, even when theres hardly any load on the GPU.

I have been scratching my head over this error for years, and the only thing I can change now is temps, everything else has been played with and tweaked. This error is literaly like a parasite, it latches on and won't let go.


Edit: I've decided to buy the xblade case and pull out some exessively high CFM fans, the fans for the xblade are 120mm for the front intake and rear exhaust, and an 80mm fan for the side. I'm going to put my industrial cooling fan on the side and hopefuly I will be able to find 120mm fans to my taste, the stock ones glow blue, but I'm not buying them with the case. And I always found glowing fans to be just a waste of power.

BUMP

I switched everything over to the new case, and now everything runs just or well under maximum recommended temps at 100% load.

But I still get the Loop, which has me quite puzzled.

The only things left that have not been replaced are the Power supply, and CPU. However, the CPU passes all stress tests on its own, so it most likely isn't that.

Now I realize the power supply has been subjected to temps beyond what it should have, so I am replacing that regardless. But the graphics card is once again my number one suspect, and here is why.

I just got a second monitor, for a simple test I hooked it up as a secondary and loaded SoftTH set up for two monitors. I loaded up LFS with nothing else running, and at idle temps.

The second I put LFS into a resolution that switched it onto both monitors the computer locked up, 3d or not. This leads me to belive there is something wrong with the card itself, but I'm not sure.

I'm replacing it regardless now, along with the power supply.

Hi all,

I got recently the same problem with a HP 8510W with a 1920*1200 screen.
I played many times on this machine with no problems. The LCD screen was broken (many times...). The technician changed first the mother board, and after the LCD screen and the graphic card (quadro fx 570m).
The first thing I tried was LFS. Bingo ! blue screen and infinite loop.
(Event ID 108 Source nv)
he driver nv4_disp for the display device \Device\Video0 got stuck in an infinite loop. This usually indicates a problem with the device itself or with the device driver programming the hardware incorrectly. Please check with your hardware device vendor for any driver updates.

I've tried all the drivers i found. Nothing works.
I got a new HD. Reinstall everything. Nothings works.
I'm "desperate".

The game worked well BEFORE the change of all the components
I will try a new request with HP.
Stay tuned !

Bon courage !

The thing that finaly did it for me was changing the graphics card, I was sure it couldent be that since it did it with two of them.

But now that I think about it, The first card, A Radeon 9000 AIW, was abused via overclocking to the point of artifacting, and the Sapphire Radeon HD2600 was just plain defective.

I switched up to the fastest card made for AGP so far, and probably the last that will be, a Sapphire Radeon HD3850 GDDR3 512mb.

I also added a new CPU cooler and sound card to make this computer last me for a while longer.

My temps are now 40ºc case and 43ºc CPU at idle and 41ºc case and 56ºc under load. This means case temp is back up to what it was before I switched cases, but thats because the new cooler disipates so much heat, and the graphics card is much larger and generates alot more heat.

The cooler I added is absolutely massive, one of these: http://www.scythe-usa.com/prod ... /040/scmg2000_detail.html

I have a very big case and this still takes up alot of space, but I would recommend this thing to anyone building a computer.

That doesn't change the fact that the error could be caused by anything. Bad ram, low voltage from a failing power supply, a bad motherboard, defective graphics card, or even a bad bios version.

I practiacaly covered that entire list with my rebuild. It finaly went away when I practiacly built a new computer around my CPU.