Print

Add to Reputation:  

What do you think of this post?

I agree I disagree  
Comment:

[Yes! I am Invincible!]

OK i was while ago talking about the ''GPU baking'' thing and the bluescreens that happened to my MSI nVidia GeForce 8800 GTS 512 OC, but the problem isn't over. The GPU xploded again but there's a thing that makes me think. The problem isn't overheating. i monitored GPU temp when playing Crysis at max. 60 degrees.
No driver problem, no software problem. here's the deal: I can play Crysis at max without probs as long as i want.
and any other steam game, but GoldenEye: Source and Black Mesa: Source are different. i can play them for a week with daily usage (1-2 hrs a day) and BAOM! Bluescreen comes up saying: HARDWARE MALFUNCTION! NML MEMORY PARITY ERROR CONTACT TO HARDWARE SUPPORT. I checked my RAM. no major problems.

Now i'm starting to suspect the game(s). Maybe the games uses too much VRAM because and it isn't optimized well enough? Somehow this thing has to do between GE:S (or BM:S) and VRAM.
Please help me with this problem. i work on a IT business and even i won't figure WHAT THE FUCK IS GOING ON?!!! 

[killermonkey]

I'll say it once and only once: Software cannot cause Blue Screen of Deaths (unless it trips DEP).

YOU CANNOT CAUSE A BLUE SCREEN OF DEATH WITH NORMAL SOFTWARE.

BSOD is caused because Windows has decided that in order to save the computer, and/or prevent further corruption, it HAS TO STOP everything. They are caused 99.9% of the time because of faulty hardware or poorly behaving drivers.

http://en.wikipedia.org/wiki/Blue_Screen_of_Death

If you post the error number (usually something like 0x0000000A) and the extra data numbers (the numbers near the bottom that look like 0x0000000 with four of them across), then we can probably attempt to help you.

Code: [Select]

EXAMPLE:

*** STOP: 0x000000EA (0x882ECA78, 0x885DDA20, 0xBA4EBCBC, 0x00000001) 

It also doesn't hurt to upload the memory dump that windows spits out: http://www.sevenforums.com/tutorials/174459-dump-files-configure-windows-create-bsod.html

[Yes! I am Invincible!]

haha not possible to send Memory dump. The system halts without saving ANYTHING. it's basically not a BSOD, i admit, it IS A FATAL SYSTEM MALFUNCTION NOT A SOFTWARE ERROR!. This is more serious problem than you think. *facepalm*

[kraid]

If it were an issue with the mod or even BMS, don't you think there would be more ppl reporting this?
Since you allready had issues with your hardware before, i bet the error is on your end.

[Proxie]

Put your video card back in the oven, it's still raw in the middle.

[Kratos]

Put your video card back in the oven, it's still raw in the middle.

LOL  oh man my stomach!

and cook us some bacon in the meantime!

[Mangley]

Have you actually tried reseating your GPU? You'd be surprised how simply disconnecting and reconnecting stuff to your motherboard can remmedy faults.

BMS and GE:S push source to its limits and do things that it was never intended to do in the first place. In our case it's our use of high resolution normal maps combined with liberal use of dynamic lighting effects that seems to eat frames for breakfast. Even though source is a lightweight engine that can run on practically anything, Valve never pushed it as far as we did (at least the version modders are restricted to) and they never optimized or fixed things that weren't going to be utilized by them for any specific title. Even now, GE:S is still more graphically detailed than CS:GO is technically speaking, and we don't even have the luxury of dynamic world lighting maps.

Try turning off dynamic explosion lights in the options that tends to resolve most of the issues people have.

[Yes! I am Invincible!]

Ok i ''think'' i fixed this. I monitored Black Mesa: Source VRAM usage and it used 480 MHz of ram. well i overclocked RAM to 1000 MHz and now there's no errors.....YET!

So the problem was that GE:S and BM:S used too much VRAM (normal RAM clock was 512 MHz)
and now it has some extra space and doesn't stuff itself full of shit.

haha, nice one Proxy xD, That's Funny

[kraid]

Not sure if it's a good idea to overclock it that much when you allready had to fix it with baking.

[killermonkey]

RAM usage is not measured in MHz. What you are talking about is the data delivery clock signal, basically how fast data is requested and delivered to/from the processor.

By overclocking by nearly 100% you can nearly guarantee to see errors down the road. But we shall see, it could be that your ram was ubderclocked for some reason and was causing the corruption issues (not common).

[Yes! I am Invincible!]

Oh sorry, silly me.  i made a mistake again. YES, VRAM is measured in MBs, not MHz, totally different thing.
yeah Memory was underclocked or it just didn't have enought power to handle all that shit, but now it can....EASILY! No problems. benchmarked GPU, no errors, temperature stays at full load at 51-60 degrees celsius, with 65 % fan speed and auto adjust. EVERYTHING IS SOLVED