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

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

So far the real reason is unknown, bad batches, frail platform, immature technology, power supply requirements? Who knows.

So I decided to jump start this thread in order to see the development of these cpu's and hope for a better xtreme-OC friendly 32nm silicon.

Please share your death experiences here and I will do my best to update the first post and summarize the experiences. Did you kill a 32nm i5/i7 with high vcore/vtt/pll?

What vtt, pll and vcore did you use? For how long? How was the death symptoms?

Thanks!


EDIT:

USER / BOARD / CPU / VCORE / VTT / PLL / POST ERROR CODE / LIFETIME
-----------------------------------------------------------------------------------

Jor3lBR - EVGA 659 - 1.5vcore - 1.4vtt - 1.80pll - 34 + single beeps - 10secs

PcCI2iminal - Tpower i55 - 1.87vcore - 1.42vtt - 1.35pll

Sniper - Giga H55M-UD2

Leeghoofd -

hiwa - E658 / 670 /1.80 /1.65/1.675/00/ 5h ln2 @ -180

hiwa - E658 /670/2.02 /1.65/1.80/00/10h @-180

hiwa - p55-Ud6 2 X 670/1.50/1.45/1.80/00/30min @10°c

rybass - p55-UD6 3x i5 670

Stefan Payne - ASUS Maximus 3 /750/ Default Clock, VCore and so on

Brian y. 3 x i5 670 1.88v / 1.85v / stock / evga e650

dpa i5 670

Ajaidev i5 661

Deux i5 650 L942B799. 1.53v i5 661 L942691 evga e647

BustaH 1x660 1x 661 and 2x 670 - evga e647

thebanik - E661 10h - Gigabyte UD4P

Here's mine:

http://hwbot.org/article/news/the_so...eing_clarkdale

Died after 3 days of use and only 20 secs of 1.5vcore/1.4vtt/1.80pll under water 20oC.

Here's PcCI2iminal's: http://www.xtremesystems.org/forums/...6&postcount=33

http://img38.imageshack.us/img38/3272/006zk.jpg

It's also known that between Hicookie and Nick they are on their 10th i5 670 due to premature death.

http://www.hwbot.org/community/submi...in_34sec_875ms
Ran 1.6v+ vcore on air with this one, still going strong :)
Vtt was 1.39 I think.

A really bad cpu, if this is all Intel can achieve with 32nm I'm dissapointed.

One 661 went down here today :( Will pm further info via the real owner.. (no idea bout NDA's and co)

Quote:

---

Originally Posted by Sniper

http://www.hwbot.org/community/submi...in_34sec_875ms
Ran 1.6v+ vcore on air with this one, still going strong :)
Vtt was 1.39 I think.

A really bad cpu, if this is all Intel can achieve with 32nm I'm dissapointed.

---

Quote:

---

Originally Posted by Leeghoofd

One 661 went down here today :(

---

May they rest in piece. ;)

Could you guys share how they died so I can update post #1? Vcore/Vtt/Pll/Board/post code #?

maybe add batch numbers to see if there's a trend ?

Quote:

---

Originally Posted by Jor3lBR

It's also known that between Hicookie and Nick they are on their 10th i5 670 due to premature death.

---

They also like to shove 2.1v down their poor little 32nm throats though:D

Quote:

---

Originally Posted by Jor3lBR

Here's mine:

http://hwbot.org/article/news/the_so...eing_clarkdale

Died after 3 days of use and only 20 secs of 1.5vcore/1.4vtt/1.80pll under water 20oC.

---

Sorry to hear that mate :(

one of my 660 cpus died after 50 seconds under ln2 pot.. 1.95v, 1.5v vtt.. temp at +13C.

this chip must have been VERY fragile.. my 670 havent died yet, and I have tortured it with 2v on air for over 3 days, still going strong :)

I think its some chips that have some problems.

Lmao, that chip doesn't want to die. Never seen anyone fail at killing a chip before, totally hilarious:D I think a torch or a hammer is what it takes...

Quote:

---

Originally Posted by dpa

one of my 660 cpus died after 50 seconds under ln2 pot.. 1.95v, 1.5v vtt.. temp at +13C.

this chip must have been VERY fragile.. my 670 havent died yet, and I have tortured it with 2v on air for over 3 days, still going strong :)

I think its some chips that have some problems.

---

That's awesome heh, surprised it's still alive after that punishment. Last time I used 2v+ on air was an old northwood.

BOARD / CPU / VCORE / VTT / PLL / POST ERROR CODE / LIFETIME
E658 / 670 /1.80 /1.65/1.675/00/ 5h ln2 @ -180
E658 /670/2.02 /1.65/1.80/00/10h @-180
p55-Ud6 2 X 670/1.50/1.45/1.80/00/30min @10°c


2 thinks killed my chips
high vtt (higher than 1.60 i think) Evga old bioses for high qpilink needed high vtt , low temp on retail chips like -175 /-180 under Wprime load (chip died when i ran Wprime 1024 after 5 seconds )

Who was running a coldbug/boot modified mainboard?

That is the 1 million dollar question...

It's quite an important question ... I have a bad feeling ...

Quote:

---

Originally Posted by massman

Who was running a coldbug/boot modified mainboard?

---

Mine was without any mods, stock board installing OSes :shrug:

mines was all with mod

Lost a gulftown.

Q3QP... it was an ES tho.

Ran @ 1.425vcore Vtt @ 1.4 on a eVGA Classy. E759

Lasted about 2 weeks... Thank god i have a X980 that replaced it tho.

Going to try a different retail this weekend. Wish me luck!

Quote:

---

Originally Posted by Jor3lBR

Going to try a different retail this weekend. Wish me luck!

---

good luck my friend :up::up::up:

3 of mine is killed...
1 is ok...i was overclocking at more than 1.9V and 6400 Mhz....and death...simply stops to work....
but the 2 another %#^~>%& death at start...stock voltages and frequency...
Its really a fragile chip
Lots of Clarkdale is dying
Good lucky with the next chip...i almost give up clarkdale

man this is scaring the hell out of me now,well this + the socket burn thread.
got an i3-530 and the H55M-UD2H w/the cheap foxconn socket and lockdown
clamp, and now im so nervous about pushing it past 21x200@4200 w/1.24v
oh well,i'll push it anyways."volts be damned":D

Quote:

---

Originally Posted by radaja

man this is scaring the hell out of me now,well this + the socket burn thread.
got an i3-530 and the H55M-UD2H w/the cheap foxconn socket and lockdown
clamp, and now im so nervous about pushing it past 21x200@4200 w/1.24v
oh well,i'll push it anyways."volts be damned":D

---

As long as it's retail radaja Intel or store has to back it up on the rma.

Problem is good ES cpu's failing are a big $300 hole in the pocket! :down:

thanks.thats whats i thought.
sorry about your guy's dead cpu's

mannnn it really suks
i was thinkin of playin wth i3
i'll wait to see if radaja kills his :p:

So far so good with my 670, been on the cascade twice and under LN2 twice with right at 1.9 vcore. Going to try again this weekend hopefully, with a different board...

Quote:

---

Originally Posted by road-runner

So far so good with my 670, been on the cascade twice and under LN2 twice with right at 1.9 vcore. Going to try again this weekend hopefully, with a different board...

---

Good to hear that road-runner my 2nd retail has the same batch as yours... Newegg?

Quote:

---

Originally Posted by Jor3lBR

Good to hear that road-runner my 2nd retail has the same batch as yours... Newegg?

---

Yes... :up: Getting my drives ready, going to throw it on the cascade tomorrow for a few hours to see how the board does and all, get it close then LN2 if I have time, if not probably Sunday... :up:

I bought another 934 batch....is the last...or it works fine or i will give up 1156...
Good lucky for you guys...waiting your results

one dead westmere-was doing 32M at 1.88v after session I ran one wprime at 1.85, when the CPU was warming up on stock bios settings at -50 and I was pulling screenshots off drive it died
One retail 670-OC station flipped out and sent 2.11v through it -150 now it autmatically corrupts BIOS in any board you put it in
One ES 670 at -190 adjusting vcore with eleet and died


That is in the past ten days or so......

OMG that's some serious losses... :eek:

Wow these things are failing left right and center. Sorry for your losses guys.

Quote:

---

Originally Posted by Jor3lBR

USER / BOARD / CPU / VCORE / VTT / PLL / POST ERROR CODE / LIFETIME
-----------------------------------------------------------------------------------

Jor3lBR - EVGA 659 - 1.5vcore - 1.4vtt - 1.80pll - 34 + single beeps - 10secs
PcCI2iminal - Tpower i55 - 1.87vcore - 1.42vtt - 1.35pll
Sniper - Giga H55M-UD2
Leeghoofd -
hiwa - E658 / 670 /1.80 /1.65/1.675/00/ 5h ln2 @ -180
hiwa - E658 /670/2.02 /1.65/1.80/00/10h @-180
hiwa - p55-Ud6 2 X 670/1.50/1.45/1.80/00/30min @10°c
rybass - p55-UD6 3x i5 670

---

Stefan Payne - ASUS Maximus 3 - Default Clock, VCore and so on, errorcode: switched off and is completely dead now, Lifetime about 3 Weeks.

@ stefan

Core i5 & i7 32nm (Westmere) Graveyard

as i know you had a i5-750?!

Ah, sorry, you're right, I had a i5/750, manufactured in 45nm.

Only one i5 has died in my hands, a 661 and that was due to voltage. A i7 980x is more tolerant than the i5 661 in my experience you really have to know how to overclock the 32nm dual cores and you cant take chances at all.

I think my all new retail 670 is "dead" also.. It just hangs in bios.. I get artifacts when I "surf" in bios setup, and then a hang.. my 661 works still!

Quote:

---

Originally Posted by dpa

I think my all new retail 670 is "dead" also.. It just hangs in bios.. I get artifacts when I "surf" in bios setup, and then a hang.. my 661 works still!

---

It doesn't looks like it's dead, otherwise you would not even boot.

Was it like this when you first tried it on the mobo? What's the batch?

I have been playing around with a new 660. So far 1.60+vcore hasn't killed it. It is a fun chip to oc but every time I go up on the voltage I get nervous.

thats good to hear.everytime i see someones name who has an i3 or i5 with a new post in this thread
i panic and think "oh hell not another one".

^ LOL , I hope to avoid adding this chip to the list. I would like to push it more but I think for now I will leave well enough alone.

It is stable at 4600mhz-1.344vcore, 1.380vtt which is better than I had expected.

Lost two chips in under two days -_-

First, i5 650 L942B799. Run briefly on SS at conservative voltages, <1.55 vCore. Ran for about tow hours before shutting down. Was working fine when I shut down, next day no post on air in multiple boards.

Second, i5 661 L942691. Only ever tested on LN2, maxed out around 1.9v around -160. Zero problems at all, until I changed vCore with Eleet on the P55 FTW E657. First voltage change and the chip died instantly, apparently I'm not the first one to observe this either.

Quote:

---

Originally Posted by Deux

Lost two chips in under two days -_-

First, i5 650 L942B799. Run briefly on SS at conservative voltages, <1.55 vCore. Ran for about tow hours before shutting down. Was working fine when I shut down, next day no post on air in multiple boards.

Second, i5 661 L942691. Only ever tested on LN2, maxed out around 1.9v around -160. Zero problems at all,7 until I changed vCore with Eleet on the P55 FTW E657. First voltage change and the chip died instantly, apparently I'm not the first one to observe this either.

---

Sorry for your lost mate.

This is a pattern happening more and more :shakes: You are the 5th or 6th person that report in with the exact same cause of death like mine.

I think the best way is to either boot from BIOS and not touch on voltages or use EVBot with nice and slow increments.

thebanik - E661 10h - Gigabyte UD4P

i took my cpu out yesterday and everything looks good so far.
the pins on the board are still in perfect condition and the pads
on the cpu itself looks great(no burns)so ill keep pushing w/out
too much fear.

1x660 1x 661 and 2x 670 no serious volts in any of them. the cold bug mod for evga e657 ftw I believe killed 660 and 670's all my chips were retail.
the online shop I use are getting very twitchy about selling me anymore:D

Had no problems at all using them in MSI or Gigabyte boards with cb mods only on the Evga boards. I have no idea what the problem is only that the mod for that board is maybe incorrect? and doing it somehow allows dirty power or overvolts the cpu? isnt this what we are all thinking?

I have a 661 still kicking and another 670 on the way. :up:

btw this happened on 2x e657 AFTER doing the cb mod, both boards worked perfectly before the mod. after the mod no post and dead cpu's.

Quote:

---

Originally Posted by BustaH

1x660 1x 661 and 2x 670 no serious volts in any of them. the cold bug mod for evga e657 ftw I believe killed 660 and 670's all my chips were retail.
the online shop I use are getting very twitchy about selling me anymore:D

Had no problems at all using them in MSI or Gigabyte boards with cb mods only on the Evga boards. I have no idea what the problem is only that the mod for that board is maybe incorrect? and doing it somehow allows dirty power or overvolts the cpu? isnt this what we are all thinking?

I have a 661 still kicking and another 670 on the way. :up:

btw this happened on 2x e657 AFTER doing the cb mod, both boards worked perfectly before the mod. after the mod no post and dead cpu's.

---

What method did you use to change the vcore vtt etc? Eleet?

both of those boards died refused to boot after doing the mod and took the cpus with them. don't get me wrong the evga boards are very good, just that mod is suspect I have no idea what it is doing, I can only guess.
The mod is very simple so not much to go wrong with the actual soldering. the boards died and killed cpu's while trying to boot on air at stock settings btw.

Quote:

---

Originally Posted by Deux

until I changed vCore with Eleet on the P55 FTW E657. First voltage change and the chip died instantly, apparently I'm not the first one to observe this either.

---

Eleet got my 670 also, sold the FTW shortly after...

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

Quote:

---

Originally Posted by splmann

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

---

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

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 !

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

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.]

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

Quote:

---

Originally Posted by dinos22

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 :p:

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 :)

Quote:

---

Originally Posted by tool_462

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 :p:

---

Definitely Agreed

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

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...

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.

Quote:

---

Originally Posted by Jor3lBR

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 :toast:

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

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 :(

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

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....:D:D

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 :)

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.:rofl:

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

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