920 lost a core..

[stevty2889]

My pentium-D 920 has apparently decided it's not a 920 anymore..device manager still shows ACPI multiproccessor PC, but only shows 1 cpu, task manager only shows 1, and CPUz now sees it as a Cedar Mill...?????????

This happened once with my 830 as well, but on a differant motherboard, with a differant hard drive, and ram. Only thing common is the video card. ??? Anyone else experienced this yet? All I did was reboot after installing folding@home as a service, and it lost a core..I have reset everything to default settings, although when it happend with my 830, it wasn't overclocking it.

 

[Furen]

does the Bios still recognize it as a dual-core CPU? I'd try reinstalling windows and disabling any overclocking just to see if that helps. Of course you could have just burned out a core but I doubt it.

 

[Cooler]

Originally posted by: stevty2889
My pentium-D 920 has apparently decided it's not a 920 anymore..device manager still shows ACPI multiproccessor PC, but only shows 1 cpu, task manager only shows 1, and CPUz now sees it as a Cedar Mill...?????????

This happened once with my 830 as well, but on a differant motherboard, with a differant hard drive, and ram. Only thing common is the video card. ??? Anyone else experienced this yet? All I did was reboot after installing folding@home as a service, and it lost a core..I have reset everything to default settings, although when it happend with my 830, it wasn't overclocking it.

did you mess with the bios i know when i was ocing my 840EE if i disabled the worng thing it would think it was reg 840 or even presccot with 3x hyper threading

Try doing a cold boot that help fix some of the stuff for me.

 

[Unkno]

update to newest bios? reset cmos? prime95 stable? temps? overclock/voltage?

 

[stevty2889]

Originally posted by: Unkno
update to newest bios? reset cmos? prime95 stable? temps? overclock/voltage?

Already using the latest bios, no newer version. 24 hours prime95 stable @3.7ghz 1.475v, water cooled, was hitting 50c max at load.

BIOS was not seeing the second core, reset bios, sees the second core again, but even safe mode is getting a BSOD, and this is back at stock speeds.

 

[thecoolnessrune]

By my calculations, the largest voltage your proc can safely take is 1.375V. Sad to say, and may still be wrong, but may have been overvolted and burned a core. Being 65nm parts, it would make sense that they would be even more sensitive to too much voltage.

 

[stevty2889]

Windows still will not boot. Tried doing a repair install, gets to the same place and reboots. Getting a stop 7F. Safe mode gets the same thing.

 

[stevty2889]

Well now disabling hyperthreading in the BIOS turns the second core off..it didn't do that before..

 

[HybridSquirrel]

good luck man. maybe you can turn it into a logging company?(cedar mill) jk jk i really hope this gets worked out for you!

 

[liebremx]

Stevty, have you cleared the CMOS? (Don't see any mention of doing that in your posts).
Try booting any of the live CD versions of Linux (Knoppix, Gnoppix, etc) just to check if it's just that your Windows installation is gone.
Also try to use explicit values in your BIOS instead of any 'auto' settings .. I own an ASUS P5WD2-P and I wasn't very happy with the guesses the BIOS was trying to make for some of the options set to 'Auto'.
It looks to me that the BIOS is reading crap from the CMOS but is not able to 'tell' that is crap for some reason (checksum not being calculated correctly maybe).
And also, as mentioned already, maybe the voltage you have used was too much: Max VID range according to spec is 1.3375V (normal operation) and absolute max is 1.55V (can still operate at this V but lifetime is degraded. Above this there's no guarantee about reliability nor functionality).. but you probably know that already
Good luck!

 

[stevty2889]

Hmm, well leaving it at stock speeds, I set the vcore to 1.4v and windows is no longer hanging..

Sounds like I may have damaged the chip.

 

[Acanthus]

That is an awfully high % increase in voltage over stock you were using... good luck.

 

[stevty2889]

The repair install ended up hanging again.

Disabled HT in the bios(which turned off the second core again), and the repair install has made in farther than it has been.

 

[stevty2889]

Repair install completed normaly with the second core disabled. Re-enabled the second core, and windows booted up fine again, CPUz showing it as a 920 again. I was installing the drivers for the onboard graphics, since I removed the PCIe video card as I thought it could be part of the problem, and I got a BSOD Machine_check_excpetion.

 

[Avalon]

Yep, sounds like the second core is damaged
Sorry to hear about that.

 

[Technonut]

I hope this is an isolated incident... It may be possible that there was a large voltage spike (mobo voltage regulator failure) when you restarted the system. Nonetheless, this is something to take note of. It makes me sad to read about it stevty2889....

Someone on another forum posted a link to an Intel Design Sheet covering the Presler. From what I gather, it appears that the processor is validated to a max Vcc of 1.6v. (Figure 2.7)

Perhaps dmens will post and comment on this validation procedure......


Edit: Fixed link....

 

[Furen]

stevty: Have you checked the chip for physical damage? Perhaps you blew a capacitor or something (The multi-chip package on these things is their newest "feature" so maybe the problem is related to it).

 

[stevty2889]

Haven't taken it out to look at it yet. I tried a fresh install of windows, and it always locks up at installing devices if I have the second core enabled. Disabling it, it runs 100% stable(so far at least). According to the data sheet I looked at for the 9xx series CPU's they should be ok up to 1.55v, so 1.475 shouldn't have been enough to kill it.

 

[BrownTown]

lol, you don't seem to be having very good luck with this chip, first you cant get it to overclock, then you finally get it owrking right and it dies... I guess you can look on the bright side, if it was a single die chip you might have lost the whole thing, at least you could sell it as a cedar mill now and get some money back

 

[stevty2889]

Well so far as long as I have the second core disabled it's not crashing at all. I may try re-seating it, and I'll check it to see if anything looks burned out. But at least 1 core is working, and even a single core 2.8 would be a big upgrade for my parents, since they are running a 1.6 northwood, so once I can afford to replace the chip and motherboard I may do that.

 

[stevty2889]

Hmm, well so far prime95 has been running on both cores for 30 minutes. Yesterday the system was always crashing in less than 3 minutes. I've got it at stock speeds, 1.375v, only other thing I changed was turning off ACPI 2.0 in the bios.

EDIT: never mind, just after I finished typing this..I heard the computer reboot..and windows locked up before loading.

 

[Unkno]

have you tried resetting cmos yet.....there has been a few suggestions to try this and never heard you say anything about it...if second core still fails, i'm pretty sure your chip is damaged then....

 

[stevty2889]

Yes, bios reset was already tried(see my second post..). Yeah, it really seems the second core is damaged. I guess there is some advantage to having 2 seperate cores for the dual core..since one died, but didn't take the other with it.

 

[ND40oz]

Have you tried it in another board yet? I wondering if it could be a bios/board issue that just came to light after some stress on the board.

 

[RichUK]

That is harsh. So it looks like it is damaged permanently?

Stevty, i think you should jump on the Yonah bandwagon, and get some OCing action out of that. When mobos and Yonah CPU's become available that is