« first day (854 days earlier)      last day (4324 days later) » 
00:00 - 16:0016:00 - 22:0023:00 - 00:00

D3C4FF
D3C4FF
23:00
@TildalWave Dude! Just make an edit that's accepted!! :D
AviD
AviD
shiiite. I think I may have burnt some cores. :S
@D3C4FF glad you happen to be here....
so I was pushing the chip a bit, couldnt get it stable at high(er) clocks. tried giving it extra vcore, lowered pll, etc... Ran a test but didnt watch the whole time, temps got up to mid/high 90's.
Now, even at lower clocks, some of the cores give errors after a while on prime95...
@D3C4FF dya think I caused permanent damage??
Thomas Pornin
Thomas Pornin
@AviD Metal atoms on the silicium substrate tend to migrate over time. They migrate faster when temperature is higher. When two adjacent wires become too close because of atom migration, electron can jump from one another due to quantum tunneling. Higher frequency rates make the CPU less robust with regards to such electrical leaks.
Or, in shorter words: it is quite possible that you caused permanent damage.
You brute.
AviD
AviD
argghh
yeah, Imma go quantum tunneling on your ass.
cant believe I killed it...
Thomas Pornin
Thomas Pornin
I once had an Alpha, rated at 233 MHz. I had overclocked it to 300 MHz.
AviD
AviD
but it is not consistent, wrt the core errors.
that is it is not always the same core.
Thomas Pornin
Thomas Pornin
23:11
After a few weeks, it began to do errors, so I downclocked to 266, then 233.
Then lower.
I brought it down to 133 MHz, and one day it ceased to work even at that low rate.
AviD
AviD
see, you should have gone for liquid cooling :D
Thomas Pornin
Thomas Pornin
@AviD Oh yeah, I used a lot of liquid cooling around that time. It was coming in pints.
AviD
AviD
well, I need practical advice now - what do I do with the chip now?
How can I check it...? Should I just let it cool off for a while, try again tomorrow at low clocks...? @D3C4FF
Thomas Pornin
Thomas Pornin
23:27
@AviD I suggest you let it cool down for tonight, and tomorrow you try it with its nominal clocking. If it fails, then you might want to return it for a defect (I don't know if your overclocking attempts will show; and the vendor might be willing to turn a blind eye on such things anyway).
Anyway, it is already late in Israel; you should go to bed.
AviD
AviD
23:50
Indeed. Thanks @ThomasPornin.
D3C4FF
D3C4FF
@AviD I thought <110 degC is acceptable range for your chip? Have you reverted back to default settings and checked what happens?
@AviD Also, told ya so :p
AviD
AviD
@D3C4FF so did I... actually 105C.
@D3C4FF hehe, exactly
D3C4FF
D3C4FF
@AviD And yes, let it cool down before you test it again
@AviD Haha, you were certainly pushing it hey? Go get a corsair H80 and whack that on top
AviD
AviD
I lowered it to "safe" OC levels, that I'd tortured for a while. Didnt reset it back to non-OC. Will do that tomorrow.
@D3C4FF no, I didnt push it to 105, thats its TJMaxx
I only pushed to mid 90's.
I think it may have peaked at 97 before I shut it down...
D3C4FF
D3C4FF
@AviD Just so your aware, the thermal sensor isn't actually the EXACT core temperature.
Its quite possible that the TRUE core temp was close to 105.
AviD
AviD
23:54
I imagine. Thats one good reason to keep it AT LEAST 10C off.
I tried to keep it around 85C, but my attention went elsewhere for a while.
D3C4FF
D3C4FF
yeah
AviD
AviD
Ironically, I was pretty much done with pushing the OCing - I had decided that it if it didnt stabilize this run, I was going to back down.
D3C4FF
D3C4FF
@AviD Ouch :( It'd suck if you really DID kill the chip then
Out of interest, at those temps/settings WAS it stable? :P
AviD
AviD
Also non-ironically, I was only pushing it for a secondary, extra-power-this-one-goes-to-eleven extra profile, knowing that for day-to-day I was going to keep it at a much more reasonable (and quiet) x43.
@D3C4FF other than the cores kicking errors, sure.
thats part of why I didnt notice it right away - the average heat was much lower, since one of them (actually I think it was a core + 1 thread) werent stressing.
00:00 - 16:0016:00 - 22:0023:00 - 00:00

« first day (854 days earlier)      last day (4324 days later) »