Windows XP: Random Lockups #2

[Solid Snake]

Okay, I've installed my system in this way :>

Formated C:\ with NTFS
Windows XP
Windows XP Update [patched with everything]
VIA 4n1's versin 4.40
nVidia Detonators 29.42 Windows XP version
Direct X 8.1 [Oddly enough it does not require me to reboot]

After all that I ran 3D Mark 2001 on 1280 X 1024 X 32, 4AA, 32-Bit Textures and 3 Buffer [basically as high as my system will let it go]. It will lock up in random places, sometimes the high polygon test and sometimes in the 'high' game tests. It is really annoying me as I haven't installed anything that tend to crash cmomputers.

However Windows XP and running programs that don't use API's like OpenGL [ such as Jedi Knight and Serious Sam 2] or DX8.1 [such as 3D Mark 2001, Unreal Tournament] will work fine and not crash. So programs like Adobe Photoshop will work fine with no troubles.

I applied AS3 with a new heatsink today, the temperatures are cooler than stock temperatures by about 7-8 degrees or so. I also applied some on my GF4, it is the ChainTech Ti4600 version and it has spring plastic holders so it automatically applies pressure to the GPU. I don't think it is caused by my GF4 over heating as there are no artifacts on the screen.

I get the ol' lock up where the computer will freeze at a frame and the sound played before will repeat it self a hundred times like ... "da-da-da-da-da-da-da-da-da ... ..."

I AM NOT overclocking. I am default speeds on everything.

My current PSU is a rated 400 Wat PowerMax and I think it is enough to power the beast, do you think? Or could it possible that I need stronger power supply unit? The voltage ranges on it aren't too bad and I have had very little problems with it in the past. I hope I do not have to buy a stronger PSU. All the is plugged into the PSU is motherboard, 5 quiet fans at 3000 rpm in total, Athlon XP +2100, GF4 Ti 4600, Sound Blaster Live!, one DVD Rom and one Hard Drive and that is all.

It does seem a little weird is that I got ALL these problems when I installed the Volcano 7, because that is when all the problems started happening again. Possible my power usage? I have a spare power supply unit, would it be wise to run the fans off that power supply unit instead (although that does requires me to put and extra hole for the power point connector).

Any ideas guys?

[Antallica]

Well this probably does not apply but when I was oc'ed at 1900+ my pc would lock up completely when winamp was running and would play a 10sec (approx) part of the song over and over. I later found that a higher vcore voltage fixed the prob.

So I'm guessing a heavy load on that psu is occuring, u seem to have a fair bit of hardware on that 400w baby.

I wouldn't be suprised if that V7+ is chewing up some extra power, mine probably does... hmm mabye I should replace that Hyena 300w ASAP!!!

[Solid Snake]

I'll try off loading some of the power to another PSU for awhile and see how that goes. I'll post some results of that..

[KingJackal]

Actually, there's an interesting point.

XP2200+ That's the top of the line CPU - up where yeilds are remarkably low. So that chip will have far less overclocking headroom. Now, I KNOW that means nothing to you - but try putting the VCore in BIOS up a bit. Say, put it up 0.1V.

Your cooling should handle it, and hopefully that'll give it enough juice. Remember - raising voltages is just the old-fashioned overclockers way to increase stability ;).

[Deviant]

Yeah try what KJ said, that may work, but I also had exact same thing with mine when I first changed to XP. Found it to be the memory, cause XP has harder memory timings on your RAM than other OS's except for maybe 2000. Anyway it wasn't my CAS latency at 2, but some other wait state that I had to change from 5 to 6 and absolutely no problems since then. So also try play with your memory timings.

[mird-OC]

have you tried removing/replacing certain components (like soundcard, NIC, video card) to see if it's a particular piece of hardware causing the problem?

it's also worth noting that people all around the world using various VIA DDR chipsets have been having all sorts of stability issues. there are a group of people tweaking away with WPCREDIT to try to improve stability... they seem to have made quite a bit of progress... i'll see if i can find the link for ya...

[Solid Snake]

So far so good. I have used two things applied from the messages above.

i) I have uped the CPU core voltage a little tinny bit (Gigabyte is annpying and it says +10% and no anything like an actual reading, oh well).

ii) Changed a memory timming from 1T to 2T. I have no idea what that is actually yet, and i intend to find out.

At the moment I have been able to run 2 full stability test [3D Mark 2001 SE to the max settings @ 1280 X 1024 X 32]. My stability test is 10 loops of each benchmark. So all in all it has been able to do 2 good hours of benchmarks with qute consistent results.

Thanks to everyone, but the story may not be over yet. But I hope it will be :) lol.

[Deviant]

Good news, keep us posted.

[Binky Stunt Cat]

i've said it before already i'm sure.....
i have NVMax which will FIX this....


and the zip wont fit....
also, currently thanks to ITS, i cant chuck it on my webpage...

[Antimatter]

Originally posted by mird-OC
it's also worth noting that people all around the world using various VIA DDR chipsets have been having all sorts of stability issues. there are a group of people tweaking away with WPCREDIT to try to improve stability... they seem to have made quite a bit of progress... i'll see if i can find the link for ya...

What chipsets are having these problems? I haven't seen any major instability issues with the Soltek boards up to 75DRV5 yet.

[Binky Stunt Cat]

its just a AMD + nVidia problem in 2k/XP as far as i know..... used to get it heaps with my Tbird 900/KT133A system...

[mird-OC]

it was most prevalent in a number of popular KT266A mobos, although i think the KT266 and KT333 are affected too... after a bit of investigation it was found that one particular KT266A mobo (i can't remember which one, i think maybe it was a soyo) was doing a lot better than all the others. so a few geeks got together and started comparing WPCREDIT registers. they came up with a list registers and the best values to use. from memory it was mainly to do with undocumented memory timings.

i can't seem to find the link right now, but overclockers.com had been following the issue for a while...

[Deviant]

Originally posted by Binky Stunt Cat
its just a AMD + nVidia problem in 2k/XP as far as i know..... used to get it heaps with my Tbird 900/KT133A system...

It also affected some Pentiums, Intel mobos and Nvidia video cards, but not as many.

[Solid Snake]

its just a AMD + nVidia problem in 2k/XP as far as i know..... used to get it heaps with my Tbird 900/KT133A system...

I have absolutely no idea what your talking about Binky. I chucked in a motherboard with an AMD [Athlon 1.4 Ghz] and a nVidia card [GeForce 3] and installed everything and it hasn't crashed since.

[Binky Stunt Cat]

It only happened while i was using 2000.... switched over halfway thru last year, before that it was 98SE and i had no troubles.

[Antimatter]

I've seen some stability problems with VIA chipsets and Win2k. SP2 seems to have more problems than SP1. The main issue was with IDE controllers but seemed to be fixed by using the IDE miniport driver.
I haven't seen as many problems with KT266A boards as there were with KT133/KT133A boards. Luckily most of the problems were resolveable.

[Binky Stunt Cat]

try this:
www.binky.orcon.net.nz/nvmax.zip
should fix it