he he oops, one Conroe core overboard!

FreeStream · 1 Jan 2007 at 23:54

Well after 3 months of 1.5V 24/7 has finally taken its toll on my E6600.

The 2nd core errors even at stock and 1.45V!

1st core works fine and overclocks like a mofo, 3.6ghz at 1.45V :eek:

But alas, I want both my cores.

So I have to RMA it, and tell the person at the other end it takes HOURS for this error to come up :/

Pitty too, didn't crash, only tested using orthos because F@H was crashing, games etc all fine :confused:

Bloody 65nm, can't stand abit of voltage

joeyjojo · 2 Jan 2007 at 00:02

You're certain it's the 1.5V which killed it?

Certain? Because this is the first evidence I've ever seen of volts killing a chip.

FreeStream · 2 Jan 2007 at 00:18

joeyjojo said:
You're certain it's the 1.5V which killed it?

Certain? Because this is the first evidence I've ever seen of volts killing a chip.

Its not dead, but it errors every few hours depending on clocks, which makes it unusable.

Im running orthos overnight at 400x8 and 1.45V see what happens, but it will probably fail

joeyjojo · 2 Jan 2007 at 00:22

That is pretty wierd. I guess you can't swap it out to check if it's something else like the mobo or psu?

Unlucky, but good luck with the RMA :/

A.N.Other · 2 Jan 2007 at 00:24

Bad news. **** happens, I guess.

Good luck with the RMA

FreeStream · 2 Jan 2007 at 00:30

joeyjojo said:
That is pretty wierd. I guess you can't swap it out to check if it's something else like the mobo or psu?

Unlucky, but good luck with the RMA :/

Nah, deffo CPU, if I use the Single Core flag in XP then it orthos's at 400x9 1.45V!! (that was only for an hour tho - I will try again tomorrow if it fails overnight)

lay-z-boy · 2 Jan 2007 at 07:21

Starting to see that sort of problem on my machine too, did 2.8ghz when i first got it, dropped to 2.75 for 100% fah stable slowly 2.7 became the limit and now i get lock ups every 5 hours at 2.6 so im back to stock.

(even then i have had 1 lock-up since returning to stock under water)

Don't know about rma, they don't make these 939 x2 4400+ any more :/

Seems my new rig might have to come sooner than planned at this rate.

Either that or my dfi expert is playing 'find the single setting that makes everything 100% stable again and makes no sense why just 1 value makes it unstable at stock'

FreeStream · 2 Jan 2007 at 09:32

Well its deffo dying

Tested at 400x8 (3.2ghz) 1.475V and it passed, problem is that It used to do that at 1.4V

Oh well.

Gonna shove 1.5V into it again and 420x8 to finally pop its clogs

mattbrown91 · 2 Jan 2007 at 09:53

Well, all good things must come to an end.

My 3000+ is still doing nicely at 2.7GHz at 1.35v

joeyjojo · 2 Jan 2007 at 09:58

lay-z-boy said:
Either that or my dfi expert is playing 'find the single setting that makes everything 100% stable again and makes no sense why just 1 value makes it unstable at stock'

Didn't know you had an expert. Great board, but so annoying sometimes. The big problem with mine has always been dividers, it hates some of them and won't boot at all, but is fine with others.

PhilthyPhil · 4 Jan 2007 at 13:29

I thought overclocking a CPU would invalidate the warranty so you can't RMA it :confused:

tom_e · 4 Jan 2007 at 13:35

PhilthyPhil said:
I thought overclocking a CPU would invalidate the warranty so you can't RMA it

VeNT · 4 Jan 2007 at 13:37

PhilthyPhil said:
I thought overclocking a CPU would invalidate the warranty so you can't RMA it

who said anything about overclocking it??

/me bundles Phil

Cuchulain · 4 Jan 2007 at 13:38

PhilthyPhil said:
I thought overclocking a CPU would invalidate the warranty so you can't RMA it

How are they going to know? It's not like you're sending back your motherboard so they can check your FSB and Vcore.

VeNT · 4 Jan 2007 at 13:39

Cuchulain said:
How are they going to know? It's not like you're sending back your motherboard so they can check your FSB and Vcore.

even then just reset the cmos!

PhilthyPhil · 4 Jan 2007 at 14:06

Well I remember back before IHSs CPUs that died from too much overclocking tended to look a bit burned, I guess they arn't going to bother taking the IHS off though. I don't know all that much about CPUs, but I would be supprised if they don't have some way of being able to tell. Maybe they just aren't all that bothered because in the scheme of things the proportion of people who overclock is pretty tiny. Good luck with the RMA anyway

divine` · 6 Jan 2007 at 23:42

PhilthyPhil said:
Well I remember back before IHSs CPUs that died from too much overclocking tended to look a bit burned, I guess they arn't going to bother taking the IHS off though. I don't know all that much about CPUs, but I would be supprised if they don't have some way of being able to tell. Maybe they just aren't all that bothered because in the scheme of things the proportion of people who overclock is pretty tiny. Good luck with the RMA anyway

Well if you literally toast the thing it would probably be fairly obvious. If the insides of it are going funny because of too many volts or whatever then there would be no easy cost effective way to tell...

Mattus · 7 Jan 2007 at 01:02

When overvolted CPUs die, it's usually from electromigration which wears away the material separating different circuits on the chip, IIRC. It's obviously on a nanometre scale so there's no easy way to tell.

SiriusB · 7 Jan 2007 at 01:44

If I RMA'd a dead-by-clocking CPU and they said "you have OCd this and now it is dead" my response would be "prove it".

Since they can't they have no choice.

Hey Lazy, 2.8 on a 4400+? Impressive. Mine wont go past 2.6GHz when the ram is at 200MHz. Well, I can get to about 2.8 but one of the cores always gives up before the other

My E6600 should do a little better. Got paid yesterday so I am sorely tempted to order now but I want my loan in my account first - dont wanna blow 600 bigguns to find the SLC have changed their mind

SiriusB