Constant Freezes with graphically-intense applications

[Nobody Special]

Not really sure if I'm in the right forum here...

For a few months now, my computer has suffered constant crashes while running graphically-intense applications; mainly video games, but also DAZStudio- a 3D Rendering software- and Zero Punctuation vids seemingly just for variety. It will freeze up completely. The screen essentially turns into a static picture- nothing moves, can't Alt-F4 or CTRL-ALT-DEL, even the mouse cursor doesn't move. If there's sound playing at the time, it will go on playing for a minute or so, then start stuttering real fast, something like a vinyl record skipping twenty times a second or so. I can't shut down at this point, the only solution is to power cycle the machine.

I've been ignoring it for awhile, partially because I didn't want to deal with the hassle of trying to fix it and potentially breaking something else and partially because it's an old machine that's seen a lot of use, and it's expected to get a little cranky. But the crashes keep getting more and more frequent, and all that power cycling can't be good. So I need a hand here.

You're going to ask me about my hardware, I expect. Well, that's problem number one- I don't know what my hardware is. This machine is a hand-me-down from a guy who got it secondhand from a guy who put it together with spare parts, and when I first got it it had drivers on it for two different graphics cards. If someone could point me towards an app that tells me what hardware I've got on my system, I'll gladly post from that, but otherwise I'm in the dark.

[Geek-9pm]

Please take a look at this post here on CH:
http://www.computerhope.com/forum/index.php?topic=20371.0
That should help. Sounds like the same problem.

[truenorth]

Nobody Special,To assist you with this problem and any future ones. It will be beneficial (indeed critical) to define more precisely the content of your computer. To do that you can download a free copy of a software program called "Everest" and scan your computer with it. It will give all the things in a list that go to make up your computer. My suspicion is an inadequate amount of RAM but that is simply a guess at this stage.truenorth

[Nobody Special]

Okay, well, I tried all that and it's still crashing. Even tried updating DirectX and cleaning out some old unused drivers, no dice. I ran Memtest 86+ in case my RAM was faulty. 5 passes over about 2 hours detected no errors, but that was the third try. The first two, it locked up after about seven minutes on test 5. So it could be the RAM, but it could also be something else. I got my hardware specs from a program called Belarc Advisor:

Operating System    
Windows XP Home Edition Service Pack 3 (build 2600)

System Model
MSI MS-7255 2.0
Enclosure Type: Desktop

Processor
3.07 gigahertz Intel Celeron D
16 kilobyte primary memory cache
512 kilobyte secondary memory cache

Main Circuit Board
Board: Micro-Star MS-7255 V2.0 2.0
Bus Clock: 133 megahertz
BIOS: American Megatrends Inc. V5.1 04/04/2007

Drives
99.96 Gigabytes Usable Hard Drive Capacity
20.63 Gigabytes Hard Drive Free Space

HL-DT-ST DVD-RAM GSA-H55N [CD-ROM drive]

WDC WD400BB-00DKA0 [Hard drive] (40.02 GB) -- drive 0, s/n WD-WCAHL4742633, rev 77.07W77, SMART Status: Healthy
WDC WD600BB-75CAA0 [Hard drive] (60.00 GB) -- drive 1, s/n WD-WMA8F2467481, rev 16.06V16, SMART Status: Healthy

Memory Modules
1024 Megabytes Installed Memory

Slot 'DIMM0' is Empty
Slot 'DIMM1' has 1024 MB

Local Drive Volumes
c: (NTFS on drive 0)    40.01 GB    9.62 GB free
e: (NTFS on drive 1)    59.95 GB    11.01 GB free

Controllers
Primary IDE Channel [Controller] (2x)
Secondary IDE Channel [Controller] (2x)
Standard Dual Channel PCI IDE Controller
VIA Bus Master IDE Controller

Display
NVIDIA GeForce 7300 SE/7200 GS [Display adapter]
DELL E771a [Monitor] (15.2"vis, s/n 8J85421B856G, January 2002)

Bus Adapters
VIA Rev 5 or later USB Universal Host Controller (4x)
VIA USB Enhanced Host Controller

Multimedia
Realtek High Definition Audio

Communications
Compact Wireless-G USB Adapter
VIA Compatable Fast Ethernet Adapter

Other Devices
Standard 101/102-Key or Microsoft Natural PS/2 Keyboard
Microsoft PS/2 Mouse
USB Root Hub (5x)

[quaxo]

Sounds like either a memory problem or overheating.

Either way, DAZ3D is going to struggle with that configuration. I tried DAZ3D out a few months ago with a slightly better spec'd system and even it had problems with it.

[Nobody Special]

Sounds like either a memory problem or overheating.

Okay, so how do I figure out which? Or is it possibly the graphics card after all? Or something else entirely?

[truenorth]

There are a number of things that can be done to narrow down the cause.You can open your case up and ensure that all your fans are running.you can clean out the interior of the computer for all the accumulation of dirt and dust that may be there. This requires attention to electrical grounding and removal of power and attention to the fans --that they aren't allowed to spin up beyond limits of operation. Those are methods of dealing with potential of overheating issues.After doing those things then you can use an available free temperature check on the computer to see if it is within limits. As far as the RAM is concerned you can use a test from a company like "crucial" to find out what your computer has and what is the max that it can take (plus what type). You can then go to the max.However given quaxo's point re his experience with one of the games causing your trouble i doubt that that will solve your problem. Not being a computer game player i cannot comment on your problems in that area other than to wonder if less demanding games cause you similar problems.Perhaps you can compare and comment on that aspect. It may simply be that you are demanding and expecting more from that computer than it is capable of. truenorth

[Nobody Special]

Okay, On the subject of possible overheating:

I found a program called SpeedFan that monitors these things. It's sending me information which, honestly, doesn't make sense. This is the readout I'm getting:

Sys Fan: 0 ROM
CPU0 Fan: 2519 RPM
Aux0 Fan: 0 RPM
CPU1 Fan: 0 RPM
Aux1 Fan: 0 RPM

GPU: 67C
System: 35C
CPU: 37C
AUX: 50C
HD0: 30C
Core: 67C
Ambient: 0C

Every temperature greater than 40 has a flame next to it, which I assume is bad. I get that. What I don't get: It says I have one working fan on the machine, but I've got the case open right now, and I can clearly see TWO operational fans- one in the back, presumably exhaust, and a big honkin' one right over the motherboard. There's a third in the power supply, which I can't see but which I can feel blowing out the back. So, why the discrepancy? Or is it a wild goose chase?

By the way, I can be 99% sure it's NOT me gaming above my system's means. It does behave like this while running Gothic 3, Witcher, and so forth, but it also does so while playing Asheron's Call, which celebrates it's 10th anniversary this year, and An Untitled Story, which is a GameMaker game.

[quaxo]

Case fans and power supply fans usually don't show in temperature/fan monitoring programs unless they're connected directly to the motherboard.

40C actually is quite good for a temperature for a system under stress. The maximum threshold for a Celeron D processor is between 67 and 69C, so you don't have much to worry about until you hit the 60 mark. You should consider replacing the thermal paste between the heat sink and processor and see if that helps any.

One thing to consider though: Celeron processors are low-end processors from Intel. They're not exactly great for intensive activities, like 3D programs and games, and with only 1GB of RAM, you will inevitably have issues.

[Nobody Special]

40C actually is quite good for a temperature for a system under stress. The maximum threshold for a Celeron D processor is between 67 and 69C, so you don't have much to worry about until you hit the 60 mark. You should consider replacing the thermal paste between the heat sink and processor and see if that helps any.

I don't know how to do that, or how much it will cost. I'm not sure if I can, either- the fan looks like it's bolted down or something. Well, not bolted perhaps, but it's not something I can undo with a screwdriver, and I'd be crazy to try and pry it up and maybe break something permanently. Besides that, if I'm interpreting these readings right, wouldn't the problem be the GPU, i.e. the video card, rather than the processor? Which would mean the heat sink there has failed?

Anyway, I've managed to cool things down by opening the case on both sides and setting up a big fan (as in, the kind you use to cool off a room) outside. This has brought the readings below 55, which is good, but I don't know... SpeedFan still marks them red, which might mean it would go lower, and I don't know if my MacGyver solution is going to complicate things worse down the road. My major worry is that the temperature will fall TOO low, causing condensation which will seriously screw things up. Any advice?

[quaxo]

Nobody Special,

My sincerest apologies. I only took a quick glance at your temperatures and read "GPU" as "CPU". I do apologize for that. 37C for a CPU temp is quite alright, like I said, they run alright up to about 67-69C. Disregard my comment about thermal paste.

As for the GPU temp being that high, that's a normal temperature for a GPU under a bit of stress. Is that when these temperatures are being taken, or when the GPU is idle?

(You might post both load and idle temps, just for reference)

[Nobody Special]

The fan is keeping the temperatures reliably below 60, but the crashes haven't stopped. Oddly, they *have* changed- the screen used to freeze up indefinately, but now it reboots after being frozen for a few minutes. Weird. But anyway, I'm guessing this means that temperature has nothing to do with my problem, so I'll just have to bite the bullet and get a new RAM stick. Now, here's the question: I'm using a MicroStar motherboard, which I understand is finicky when it comes to which brand of RAM it will accept. So, where can I find a Gigabyte that will work with the motherboard, but still be transplantable to something more common if worst comes to worst and I need an entirely new machine?

[truenorth]

Nobody Special, Once you have decided what the total capacity of your RAM for that computer is (Crucial will run a scan that will tell you that). Then it becomes an option for you. You can add to the ram that is there to bring it to the max.In which case the best suggestion is to match exactly the brand and type that is already installed.The 2ND option is to purchase either as one stick the total of the ram allowed and then type becomes the most important concern. If however you want to achieve your max by the installation of 2 or more sticks (depending on available space) then you need to purchase all the same type and brand. Normally type is the real factor and brand is less so but to be on the safe side i reccommend all the same type and brand. Crucial and Kingston are 2 reputable brands (there may well be others). truenorth

[Spoiler]

Could I suggest that you unplug your RAM and move it to the slot closest to your CPU?

[Nobody Special]

Fixed. It was the graphics card. I figured this out when, on a hunch, I took it out and reverted to on-board video. The crashes stopped. So I ditched my old card and got a new one, and now everything's working fine.

I suspected it was the video card in the first place, because the problem only ever happened where heavy graphics or video was involved. And I note, with some annoyance, that I brought up TWICE in this thread the idea that my video card was the problem and was both times redirected to RAM and overheating instead. Thanks for the help, CH. You stay classy.