Talking Core i5 & i7 32nm Graveyard - share your RIP's

[splmann]

1x i5 661
1x i5 750
1x i5 670ES L934xxxx

[Jor3lBR]

1x i5 661
1x i5 750
1x i5 670ES L934xxxx

Please share how it occurred. Eleet? Vcore? What bench you were running (spi, wprime)?

[splmann]

i only know it from today my 670es died after 30min. on a Gigabyte Ud7 cb-cbb mod and a special vcore bios . Vcore was around 1.9 to 2.0 Volts @ - 170. Bench was 1m Pi !

[Musho]

Weird that these processors are dropping like flies when put under xtreme benches, yet the 980X seem fine, eventhough they are made at the same node. Hmmm :S

[BustaH]

does anyone think that any of this could be to do with the integrated graphics in the chips? I ask because my 661 that died would not boot in ud7/ gd-80 but did boot to bios on a h57 board. last board is cpu ig capable first two are not?

[my issue with cb mod is unrelated to this btw.]

[dinos22]

What happened to Intel? They used to make indestructible CPUs.....

[tool_462]

What happened to Intel? They used to make indestructible CPUs.....

I think fragility is a byproduct of the process shrink. It seems they are getting "killed" more than "dying" if you get what I'm saying.

But yeah, 45nm Core 2 chips and especially 65nm stuff was more or less indestructible

[dinos22]

first chips i killed were 45nm and it hasnt improved you are right it might have something to do with it but it's not what we expect from Intel

[RAMdHERsoftware]

I think fragility is a byproduct of the process shrink. It seems they are getting "killed" more than "dying" if you get what I'm saying.

But yeah, 45nm Core 2 chips and especially 65nm stuff was more or less indestructible

Definitely Agreed

[gurusan]

Sort of odd that the VID range is up to 1.4V as opposed to the 1.3625V on the wolfdales....

[Falkentyne]

Wrong.
Absolute max on wolf/yorkfield was 1.45v.. Absolute max on 32nm (at least on the 980X, don't know about the other ones) is 1.4v. I think the absolute max on the 45nm core i7 was 1.55v. The functional VID range for the 45nm Core2's was something like 0.875v-1.3625v.

Please do NOT confuse "absolute max" with "safe limits" or "Functional vid range". Seems like some people think that absolute max means "safe voltages for 24/7 operation" for some reason...

Absolute max (voltages above the normal safe functioning range) was designed to deal with voltage SPIKES where the vcore could very momentarily jump above the max safe voltage range (aka transient voltage). This is what "Vdroop" is supposed to counteract, by the way...

[gurusan]

Nothing I said was wrong, I'm fully aware of the difference between absolute maximum and functional VID range. The absolute maximum on the Clarkdales is 1.4V which is the same as the maximum functional VID for these chips...which is peculiar.

[miahallen]

i5 & i7 32nm were launched and we are all aware of how frail these 32nm cpu's can be.

Hey bro...I missed your thread and started another specific to GT...can you get the data from my OP there, and I'll let that one die? Thanks

http://www.xtremesystems.org/forums/...d.php?t=248996

[ajaidev]

First my 661 died and 1 weeks ago my 670 died and i had acquired it with a lot of difficulty


It was a L934 Q batch series and intel DH55HC with Patriot Sector 5 Edition, the reason for its dead is not known it was working ok one day and all i did was start a stress test came back to find the computer stopped responding and restarting it gave me weird blocks and artifacts cant even go till BIOS.

Will try it with a P55 and verify the problem, it maybe a problem with just the igp part i hope

[Pt1t]

one 661 34B, very good before to die ... explosed on 657FTW and board too.

Vcore 1.9V , VTT 1.35V , Vdimm 1.6V, Vpll 1.8V

one Q3FE week 28 is now Dual extreme cpu, 4Cores RIP :p

[Hazzan]

2 670 LXXXB909 - LXXXB954...EVGA E659

LXXXB909 : V core : 1.725 CPUVTT 1.35V VDIMM 1.6V PLL 1.8V

LXXXB954 : V core : 1.8V CPUVTT 1.375V VDIMM 1.6V PLL 1.8V

1 i5 750 LXXXC304 on MSI GD65.. Vcore 1.575V CPU VTT : 1.425 Volt PLL : AUTO

All motherboard modified ..under LN2....

[Toolius]

1x661 - 1.6V Core - 1.35VTT - 1.65VDIMM - 1.84V PLL - Gigabyte P55A-UD4P - LN2 - 1 Hour

Other i5 650 and 661 still going strong after several sessions

[BustaH]

1x Q3FE working only 5 cores [10 threads] im not sure but I think it was either after I raised vtt or nb a notch trying to clock higher.
strangely it will boot to bios on 6 cores [12 threads] but restarts on windows loading, no amount of voltage tweaking will boot to windows. but it works perfectly with 5 cores [10 threads] and handles the punishment well. so im kinda happy with my quint core cpu.

forgot to add I also have a es 670 that wont boot to bios unless it has 1.4vcore, works perfectly otherwise.

[MaJ0r]

1 661 ES. I didn't kill it, but cpu didn't load OS with cpu voltage less 1.25v. It means one things - this chip will never work in auto mode. This happened after 2 session with LN2, vcore -1.92, vtt around 1.4v