Has anyone actually killed a core 2 quad w/vcore+temp?

[charleswrivers]

Now that the core 2 quads have some age and many people have had them come and go for newer stuff... for those of us that still have things like q6600 and have adhered to not exceeding recommended vcore or temp (1.5 and 72c for me) ... I'm curious, has anyone actually killed these chips? I'm pushing 3 years on this chip being at 1.4-ish vcore and regularly being in the 60-65c range at 3.4.

I'm at 1.5v and 3.6, tweaking a little for better stability (I still get a rounding error here and there) and hitting 65-70 and, well... honestly... I don't care about longevity that much anymore. I plan on getting an ivy bridge in the next year or a sandy bridge if I kill this. I can go liquid cooling to reign in temps if need be... but even then... eh. So what if I'm crossing 72c.

So lets hear it. What's the worst anyones ever done to a 775 LGA board or core 2 duo/quad? I've looked around and can't find a horror story yet...

 

[charleswrivers]

Y'know... I just went to yahoo and tried a few different things trying to find something on the internet concerning someone killing a chip.

I didn't find anything. I know you could probably kill a chip if you wanted to... but without specifically trying to... even if half a decade... has anyone actually done it?

This question also extends to people who bought legacy stuff build for a LGA 775 socket. Ever buy old stuff that was DOA or died soon afterwards. I'm really interested in the longevity of a whole generation of products that, despite what we've thrown at them, I've found no evidence they've ever (reasonably) failed.

 

[charleswrivers]

After looking at Overclock.net... looking specifically at q6600s since they're likely the most common 65nm cpus and what I'm most interested in since it's what I still have... nope. Don't see anything. A few board issues, almost all that were fixed w/clearing the cmos. A psu not being up to snuff... a hangup w/one of the cores temp reading that fixed itself. No dead cpus.

Looks like there was a 2500k or 2600k that they think they put 1.8v to and might have died (didn't think you could put that much vcore on one... maybe they were looking at their ram voltage) and I didn't read to it's conclusion but I'd definately call that a 'try to kill'.

I've seen things in the 1.7+ range for benchmarking *which obviously would have been very short term) and quite a few posts claiming running 24/7 in the 1.6 +/- .05 range. They never seem to have come back saying they killed their chip, when they upgraded.

 

[charleswrivers]

Oh well. Just ordered an Antec 900 v3 case to replace my chopped up, extra fanned uber cheap, when I got it, Xion case, and a little bit better cooler than the Artic Freezer I have now and see what it actually takes to kill this chip. Newegg had Xigmatek DKs for a cool $20 plus a card reader. Mine's SD slot died so... well... it's worthless. I can keep the case for the next build.

Like a friend's Corolla when I was in highschool... no matter how hard I beat on it... I bet I can't kill it. I think I'll just hit a wall of stability that no matter how much voltage I throw at it won't help until I get thermal throttling or a system shutdown by hitting something in the realm of 100c. The consolation prize will be having to settle for Sandy Bridge rather than Ivy Bridge. Boo-hoo.

I bet the 780i will give before the Q6600. I have a spare 680i tucked away for donor caps if need be or just to limp along at 3.4 ghz (the best overclock I got on it) if the mobo dies.

 

[jsc]

I am still running Core2 systems:
GA-EP45-UD3P | Q9550 OC'd to 3.6 GHz (425 MHz X 8.5) C3 stepping
GA-EP45-UD3L | Q6600 OC'd to 3.6 GHz (400 MHz X 9)
GA-EP35-DS3P | E7500 OC'd to 4.1 GHz (373 MHz X 11)
GA-G41M-ES2L | E6500 OC'd to 3.87 GHz (352 MHz X 11) G41 FSB limited

They are all operating well within the recommended Intel vcore/thermal envelope.

 

[charleswrivers]

Well... just put everything back together. I used to run 67-75 on my cores on prime w/the old case and the Artic Freezer w/it's push pins. Not I'm 47-60 through the cores after 15 min on prime w/that Xigmatek and it's screw in design and bigger fan. Hopefully the deltas will come down a bit once the AS5 cures a bit. Funny... my old hottest core is now the coolest. It was also the one that failed w/rounding errors. Not a rounding error yet.

After 4 years, I really wasn't very impressed w/the Freezer heatsink. Sure... it was better than OEM (what isn't?) but I had to swap out the push pins once because they got loose a year or so ago and I think they, again, were loosening up.

Anyways... time to play...

 

[charleswrivers]

Now 3.7 prime stable for a few cycles @ 1.55 bios(1.53 actual). gtlrefs 85-90-85-90. Going 90/90 on 1 and 3 got me from getting a rounding error in <10 secs to not one through a few cycles. I'm going to have to spend a while dialing some stuff in. It's fine for gaming like this but I'd sure like to see those rounding errors go away w/o having to fling >1.55v at it from where I'm at now. SB is maxed out. Think I may have to go higher on the NB. Still at 1.2 which was supposed to be good out to 400 on fsb but now that I'm going past that it may just not be cutting it anymore. SB just won't hold a heavy overclock unless it's near or at max.

To those using and artic freezer 7 on an LGA 775 board... honestly... when I got it years ago, it was obviously better than stock but now that I put that Xigmatek DK on it... it is SERIOUSLY cooler (by about 5-7c on the hottest core) and MUCH quieter. I'll never get a heatsink w/a 92mm fan again. The 120mm move air better at lower RPMS.

Oh well. You live and learn. I've been putting maybe a $100 a year over the last 4 1/2 years selling old stuff and replacing it. E6600 gave way to a Q6600. 8800 GTS 640 (the G92s weren't out yet) to a 260 gtx. Then... got an extra recertified 260 gtx for SLI for a hundred bucks. The sold both and replaced for 460 gtxs for about $50 out of pocket once I sold the game codes (HAWX 2 was god awful anyway). Swapped the 680i out for a 780i for, like $20. The 460 gtxs themselves are at 880 core. Great cards. My power consumption actually dropped going from the 260s to the 460s, not to mention about a 20-30% gain in FPS stock clock to stock clock. The 260s were space heaters.

LGA 775 has been a great platform. I'll (almost) be sad to put it out to pasture. I had a new mobo, 2500k and DDR3 in a newegg shopping cart and just didn't want ot part ways w/what I have left until I've run it into the ground completely.

Well... I'll get her dialed a but better for stability and shoot for 3.8 next. Doubt I'll ever hit 4.0, certainly not on air... but I'm only at 66c on my hottest core in prime as of now, when I was going in the mid-70s on the AC7. I haven't even really burned the AS5 in yet...

Kind'a fun anyways. I've sat around w/3.4 for a long time. It'll be neat to see what I can get out of the old chip before it gives up the ghost or IB finally comes out.

 

[charleswrivers]

Ha! Solved my rounding errors...

Never thought I'd have success going w/a higher fsb and dropping my multipler, but 475x8 is working beautifully and prime keeps on chugging away. I'm at 1.52 vcore and need to drop it down to see what I start to lack stability in which should improve temps (range from low 70s to mid 60s in blend, obviously small ffts would be... 'toasty').

I've booted at 4.05 Ghz a few days back but would BSOD if I tried to do... well... anything except move my mouse (and slowly at that).

Still, I'm very amazed at what I've managed to do... and all the while not having to really exceed 1.55 vcore (might even get south of 1.5 again. 1.2625 vid chip... so good, but nothing insane.

Again... after all the looking I've done, I've still yet to find a core 2 quad die. Seriously. Managed to find a few mobos eletrolytic caps go. Heck I had to solder 3 or 4 in my Samsung TV a while back. These were rather... entertaining though.

Highest volts I've seen is 1.9 thus far. Lots of 1.6-1.7 for hardcore guys I'm guessing kept their systems for 1-2 years tops. Many of those guys were in the 3.8-4.0 range in the process. Lots of watercooling too. I'm just cooking eggs on my chip but it sure is stable as a rock as of now.

 

[Designs]

hey any updates on the "kill a Q" thread? i recently did a lot of research on Core 2 deaths and same as you came up with nothing.

yet with the old Q6700 i was always worried about temps even though i use to push 75C + per core with "fun times" managed a 4Ghz boot once but yea BSOD, it still works even after years of abuse

 

[charleswrivers]

Nope!

Found that some 45nm C2Qs were dying due to high VTT (I think once 1.45v was exceeding) when OCing runs were being done. Theres a few warning pages a couple years old now floating around, though the 65nm weren't suceptible to dying there. Can't remember when our max VTT was but I imagine it's at least 1.5. Don't think anything much higher than 1.3-1.4 is needed anyways. I run 1.3 vice 1.2 because I'm at 475 fsb. Found that, for whatever reason... I'm hours on prime stable at 475x8 vice trying to go over 400x9 to crack 3.6 Ghz.

But... really... no... nothing. Small FFTs have me touching the mid 70s on air. I'm unconcerned. If it dies... it dies, but I'll be surprised if it does and the extra Ghz helps slightly w/minimum framerates and on a few games which hog up the CPU. Any... really, it's kind of nice to have a 3.8 GHZ Q6600. I'm under 1.5v (1.46 ide, 1.47 load after pencil mod)

I can boot and start running prime @ 4.0 GHZ w 1.55-1.6 but I'll start touching the 80s then. I haven't let it go very long (10s of minutes) but it doesn't BSOD and doesn't get a rounding error. If, in another year I STILL have this rig (and I may) and I have anything that my cpu seems to be limiting me, I may knock it up to 4.0 GHZ again. A 9x multiplier just, for some reason, nets me rounding errors here and there. Dunno how/why 8x multiplier and (for what I would have thought to be) an excessive fsb speed would be better.

Realistically, the mobo should give before the cpu.

I don't know. I spent a week after my last post playing around. I haven't don't much since then. I might try to break the 4.0 GHZ mark again and see what temps end up being over time... 80+c or not. Realistically, I oughta have watercooling as I'm pulling a 67% overclock... and that extra 200 Mhz is needing a whole .1v or so to hang.

So in there era of newer, faster hyperthreaded CPUs... yeah... maybe over heat and CERTAINLY overvolting may be a concern... but it seems the 65nm design was rather robust and, now that a few years have gone by, I can't find any evidence of dead chips caused by excessive heat/volts. Run it till it dies. The Q6xxx will end up being that crappy car everyone knew... that you wanted to replace and beat on it w/all your might... but it just wouldn't die, and it kept getting you from point A to point B.

 

[clonazepam]

My sandy bridge setup just arrived. The first thing I thought about doing was pushing my Q6600 to its limits. Asus P5N-D board, 750i chipset. I haven't taken it to 1.5v yet, 1.4875 is the max so far. I had it up to 3.83ghz long enough for a cpu-z valid but that was it. If you haven't already lapped the cpu, go for it. I lapped it and the corsair H50 Im using to cool it (with push/pull). I've had the Q6600 up in the mid-80s. I was biting my nails lol... Now that my new setup is here, I'm going to go above the 1.5v barrier and see what she's got. I was / am very fond of my Q6600 as well lol... but better to go out in a blaze of glory I suppose.

 

[charleswrivers]

Sounds like fun.

Again, I could crack 4.0 around 1.55v... hitting 80-ish in the spring.

Love to hear how it goes. I've seen lots of suicide runs at 1.6-1.7v that didn't ever result in a dead chip. I think you'll find just a lack of stability w/o killing it... or just have it shutoff.

Let me know how it goes. In a weird, morbid way, I'd actually be happy to hear you managed to kill one. You'd be the first post of a dead CPU I'd of found for a 65nm C2Q. I did find some 45nms go from high VTT. You might have your board go first, before you can get the CPU to give up the ghost.

I did lap the Q6600. I managed to get about a 2-3F drop. Never messed w/WC. It likely would help temps drop, but... eh... I just don't care enough anymore. As Dolph Lungren once said, "If he dies... he dies."

 

[clonazepam]

I forgot about this thread. I'll be doing some cable sleeving and tidying up of the new system next week. It'll be the perfect opportunity. I still need a super low power machine, but I can do a lot better than a Q6600, so we'll see how it fares next week.

 

[charleswrivers]

Ever kill that chip of yours?

 

[clonazepam]

Nah, not yet. I picked up a Samsung Smart TV... but almost none of my video files is compatible. Most of the time, the video plays while giving an error that the audio format isn't supported. I've been hustling and bustling to find free software to convert the files to other formats like avi, wmv, and asf so I can stream them fron the NAS, but there's either a huge loss in quality, it doesn't play at all, the video plays with no audio, or I cant get the subtitles right, lol...

So, for now, the Q6600 based system is a really big htpc. I have to build an actual htpc, either with a llano chip, or an inexpensive bulldozer product. Maybe I'll have that done by Christmas. Even after that, I may still find some use for the Q6600... I think I'm just in love with it after all the tweaks and add-ons over the last 4 years, I just can't do it. lol

 

[charleswrivers]

Oh well... at least the old chip is still finding use doing something. Mines still driving this rig. I didn't get the nvidia 5xx series because the 460 SLIs I have do anything they can do (ie DX 11) and the Q6600 @ 3.8 won't bottleneck most games. Maybe next year... but as of right now... unless a new generation of GPUs come out that actually do something new... I think I'm sticking with what I have.

 

[akataria7]

hey man . the other day my PC wont start. it was core2 E7500 . i found that the RAM wasnt inserted properly . i made a mess in my CPU nd accidently my heatsink wasnt properly installed. when i started my CPU the temps hit 78 at idle nd 117 under load :O

 

[johnners2981]

hey man . the other day my PC wont start. it was core2 E7500 . i found that the RAM wasnt inserted properly . i made a mess in my CPU nd accidently my heatsink wasnt properly installed. when i started my CPU the temps hit 78 at idle nd 117 under load :O

117 under load? highest temperature I've seen is 97C in my old laptop under load

 

[clonazepam]

Oh well... at least the old chip is still finding use doing something. Mines still driving this rig. I didn't get the nvidia 5xx series because the 460 SLIs I have do anything they can do (ie DX 11) and the Q6600 @ 3.8 won't bottleneck most games. Maybe next year... but as of right now... unless a new generation of GPUs come out that actually do something new... I think I'm sticking with what I have.

I had the Q6600 at 3.6ghz driving 2 570s in sli. It severely choaked the performance. You probably wouldn't want to pair anything higher than 460s in sli with the chip anyway. I dropped it back down to 1 570 and am auctioning the 2nd one. If it doesnt sell maybe i will do a micro atx sli build with that inexpensive bulldozer part.... we'll have to see how that all pans out.

 

[akataria7]

117 under load? highest temperature I've seen is 97C in my old laptop under load

ya it was 117C . i was totally blowed

 

[charleswrivers]

I figured a thermal shutdown would be around 100 C or so. 117 is toasty. At least you can double the cpu as a range top to boil water with if need be.

The cpu bottleneck is all based on the game. Some will definately bring a core2quad to it's knees, though usually still with goodm playable frames. It's rare to see the Q6600 touch 100% in gaming... while the 460s can regularly be topped out. 570s in SLI would likely be 1.5-2x 460s I'd guess. I know 460s SLIed will trade blows with a single 580. It's just nice to see after about half a decade that something so old is still working so well. I'd really just started this thread to see if anyone had killed a 65 nm chip or would at least try to kill one they still had lying around to see if it could be done.

It seems they're the equivalent of a friend's Toyota Corolla in high school. Not flashy. Not fast. But damned if you can blow it up, no matter how hard you try and it keeps getting you around.

Nvidias 5xx do nothing a 4xx won't already do, they just do it better. I'll ebay those when nvidia has an offering with new features. I'd also like to see something mainstream > quad core (not to say multithreading isn't cool and very effective). Perhaps when intel brings something very, very substantial over the core2quads... I'll take the plunge.