BTW, have you called your hardware vendor yet? ;-)

Nope :( My "hadrware vendor", the shop where this PC was bought (usually I build PCs from parts by myself, but this one was bought as whole for some bureaucratic reasons) cease to exist more than year ago... But anyway, PC's age more than warranty was. So I should repair it on my own :)

The times I've had sporadic hardware lockups, specially under high CPU load, it has usually been the motherboard. Hope you have better luck than I've had.

BTW, have you called your hardware vendor yet? ;-)

I got That Same BSOD a week or so ago.
After two times of that I got an ATI ERROR Message telling me that I had a driver in a loop or a hardware failure.
Got a new low end ATI (HD3470) card and will be putting it in on Saturday...Hope it Fixes My Problem.

---

I Desire Peace and Justice, Jim Scott (Mod-Ret.)

Please, look at this BSoD screenshot.


Any ideas what hardware component has problems?

Can we consider this "informative" message
as 1st April's joke from Microsoft? ;)

Where do you read voltages (some program, e.g. SIV - System Information Viewer)?
Are they fluctuating too much (especially the CPU core voltage)?

---

 


- ALF - "Find out what you don't do well ..... then don't do it!" :)
 

This is the error message Bill...

*** Hardware Malfunction
Call your hardware vendor for support
*** The system has halted ***

Please, look at this BSoD screenshot.


Any ideas what hardware component has problems?

For some reason that looks like a blue postage stamp on my screen <32 inch LG monitor/tv running at 1080i> and when I blow it up it is too blurry to read. None the less, with these kinds of intermittent problems I first blow out All the dirt from Everywhere in the case, then Reseat Everything. Loose connections or a corroded connection can do this. Then I'd see if it has problems running Linux off a cd/dvd to try to eliminate software. Then check memory. Then check power.

OBTW, thanks hugely for your coding efforts.

(D)RAM can be a troublesome thing, so are already too much components, causing the trouble eliminate 1 by 1, seems to be usefull, besides, a combination of conflickting hardware is possible too, will show up inmediatly.

In a few years, I've enough to start a computer part store? And I had some very old models and test-PC's, all stolen during a move from Deventer (Overijssel) to
Zutphen(Gelderland), truck wasn't payed attention to for 25 minutes....

(And the Police, I had to call the next morning again and my insurance man, was there, before the 2 cops arrived.)
Ehh, Back on Topic

---

Hard to tell about frequency of errors (cause I use that host mostly remotely and can notice failure after few days passed...).
I tried to lower CPU/mem voltages - errors present, tried to go to higher values, error stil presents... I think I will not put Corsair there (cause I had very hard time to fit it in quad's case), but I will try to swap that PSU with now idle old InWin PSU from quad. Good to have some spare parts :)

Sounds like a plan. At least that should isolate PSU issue or not, which is one possibility down. 9600GSO is a relatively light load, so I'm not surprised it wouldn't have much effect, especially if it's on a separate power rail to the CPU.

FWIW: It doesn't 'look' like any RAM related issue I've seen with failing sticks. Probably can't rule that out though. I hope it isn't the NorthBridge.

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

Hard to tell about frequency of errors (cause I use that host mostly remotely and can notice failure after few days passed...).
I tried to lower CPU/mem voltages - errors present, tried to go to higher values, error stil presents... I think I will not put Corsair there (cause I had very hard time to fit it in quad's case), but I will try to swap that PSU with now idle old InWin PSU from quad. Good to have some spare parts :)

YEs, I thought about bad PSU, but don't understand why host not sensitive to GPU load then. GPU consumes much more in fully loaded state....
Well, it looks like I should replace part by part until broken one will be found...

Have checked all BIOS settings?
Did you use TThrotle before? One of the progs ASUS has put on a DVD was a
program Smartdoctor, worked OK with GTX480, but not with 470?
And due to some strange 'workaround?', the same type mobo runs 480 @16x, butonly
in slot 2, slot 1 runs in 1x mode. Same story with 470.....

While 2 ATI a 5770 and 5870, both run in PCIE(2.0)@16 at 16x???

Hardware is already extremely complex and gets only more complex in the nearby
future. :)
The CPU, Northbridge Southbridge, Memory?

[ADDED]
And which type/model?

---

YEs, I thought about bad PSU, but don't understand why host not sensitive to GPU load then. GPU consumes much more in fully loaded state....
Well, it looks like I should replace part by part until broken one will be found...

That is weird when that happens, It can be one rail in the PSU gone flakey (unless it's single 12 volt rail). If you have the option you could try increase motherboard voltages 1 notch & see if the frequency of failure changes. If it decreases then likely power issue somewhere through the line from PSU to motherboard. If it increases in failure rate then likely temperature or borderline component somewhere.

If you borrow the corsair unit from the other machine for a short test period, does the failure stop ?

I think at this stage in development it's important that your NV setup is stable too, since your Ati work versus our fermi work needs solid comparison to pick the best bits for future work.

I would suggest reseat every connector & board (even the CPU), swap temporarily , one by one, with 'known good components' and if that doesn't narrow things down suspect the motherboard. A last ditch BIOS flash might not hurt. If a can of frozen air sprayed on the northbridge delays the failure longer than usual, drop all the volatges a bit.

What do decent motherboards cost in Russia ? We know about good PSUs now.

Jason

---

"Living by the wisdom of computer science doesn't sound so bad after all. And unlike most advice, it's backed up by proofs." -- Algorithms to live by: The computer science of human decisions.

YEs, I thought about bad PSU, but don't understand why host not sensitive to GPU load then. GPU consumes much more in fully loaded state....
Well, it looks like I should replace part by part until broken one will be found...

Well, you are right about cryptic but also complecated as no particular
Device is mentioned.
A PSU, at the end of his life-time/span, or even bad electrolyte capacitors, which you can sometimes see, cause they expand and can be
even explosive!
(A 100mF, 0.1F 35V, in the PSU/tranformer/rectifier, in and 500Watt Audio amp exploded and ruined all components in it's neighbourhood!)

[ADDED]
Cabels, like ethernet, USB1.1/2.0/3.0, eSATA, Firewire.
Cables from PSU, fans, [i]SoftWare HDD's, especially when using RAID (AHCI).
DVD/CD/BlueRay.........
I need a rest.........;^)

---

Really a cryptic message, which I've seen never before and I've
seen quite a lot

When does it occur, after BOOT POST and Loading Windows?

It "just happens". In rare cases on boot stage, mostly just happens in operational stage.

The most interesting part: it's Windows Server 2003 x64 OS, but when I installed Windows 7 x64... I met just THE SAME message :)
No failed driver mentioned, no any debug info... very cryptic indeed....

Oh great, are we going to have to get you a new computer to go with your new card? :-)


Please don't miss the smiley, I'm only joking!

LoL, maybe NV part now ;) Cause it's NV-enabled host. I use it remotely most of time and these sporadic BSoDs slow down NV AP/MB development a lot :(

Really a cryptic message, which I've seen never before and I've
seen quite a lot

When does it occur, after BOOT POST and Loading Windows?

DRAM sticks or PSU, since you do have a visual, your 8600GSO is OK :)

(Sometime ago I got similar Hard Ware Failiar, a PCIE@1x used by a
GigaBit Ethernet card.)

---

Oh great, are we going to have to get you a new computer to go with your new card? :-)


Please don't miss the smiley, I'm only joking!

---

PROUD MEMBER OF Team Starfire World BOINC

If the machine will start, maybe in Safe Mode if nothing else, there might be an event near this in the event log that would help find the issue.

The issue with that is this BSOD happens during boot. Maybe he could cut it off before it happens by going into safe mode. Raistmer I would never question your coding abilities, but you don't think you may have killed the card with some new ATi apps did you?

No :) Maybe I killed that host by installing both HD4870 + GSO9600 in it some time ago, but failures was before too.
It's not on booting time. It happens time to time, sometimes very often, sometimes few days pass...

What I know for now - if both CPU cores (it's Duo) loaded - host will hang very soon. So I run it with only 50% setting in BOINC. GPU load seems have no effect on host stability.
There is no direct relation with CPU temperature, cause I use TThrottle to monitor/limit temp. Sometimes it can sustain rather high temps, sometimes it gives BSoD at quite cold CPU...

If the machine will start, maybe in Safe Mode if nothing else, there might be an event near this in the event log that would help find the issue.

Found nothing in event log, I tried few times....

Could be anything. This is what Microsoft says about it.

http://support.microsoft.com/kb/q222973/

Mainly they say it could be memory, adapter cards, or a bios setting set wrong. My first guess would be a failed bit of ram, especially considering you have a new video card. Of course it could be that, a failing power supply can also cause these issues.

No, it's not my quad with new GPU, it's another host with GSO9600...