Have you fried a chip when you were sure you were safe? Post how here.

[soccerballtux]

Something BD321 or whoever said caught my attention and I've been turning it over in my head for the past week. He was running a ph2 955 at 1.45v and it went belly up on him.
I didn't investigate further but thought I'd go ahead and make a thread about it. I am cheifly interested in AMD chips (Phenom 2) but don't let that stop you, this is about whatever CPU you killed.

Myself, I have a Ph2 x3 720BE sitting around that had an intermittent issue with suspend since the day I bought it, I just never tested it thoroughly because I was sure it was thanks to unlocking it to a quad + overclocking. Well, after 2 years of overclocking moderately with it (see sig) 3.5ghz/2.6ghz @ 1.475v and 1.4v CPU-NB/L3, I finally was able to reproduce the bug with consistency. Heat it up past about 50C, suspend machine, wake it back up, and then the L3 cache is flakey until a full power cycle. IE Reboot not good enough, must turn off, unplug from wall/turn off PSU. Very odd. Debating whether it would be ethical to warranty it considering I am now certain the problem existed from the beginning thoughts? Anyways, so, I upgraded to a 965 and am running at 4ghz/2.6ghz 1.475vcore/1.35v L3, happily, because 95% of the time the cores are in their sleepy state running at a peaceful 1.07v. CPU-NB I have found does not ever come down from 1.35v though...a little worried about this because I do want to crank it up a little faster, I seem to be having instabilities here and there (L3 cache problem goes away when I turn voltage to 1.2v and clock it at 2.4ghz, but 2.4->2.6ghz is noticeably; I will not give it up 🙂).

But that is not why you are here. You are here to be a part of a NEW ERA in forum collaboration. I am looking to make this thread a thinktank of damage-cpu-related-information. Never ever done before, so many posts on one html page, wow!! think of the possibilities!!! Continue below--


TL;DR--
1. Have you killed CPU before? No? Why aren't you overclocking harder then??
2. If you answered yes to #1, please post the CPU architecture, frequency(s), voltage(s) that you killed it at
3. Give your best guess as to how you killed it.
4. Were you surprised when it died? Were you overclocking le hardcore while folding 24/7? Well less of a surprise then right? In contrast, BD321 was surprised because he was within AMD's own recommended max voltage (1.5v, he was at 1.45) and his chip died.

TL;DR version of TL;DR version:
did you kill a cpu why were you surprised how did you kill it would you do it again

thanks to all, and to all (especially our Native American friends 🙁🙁🙁) a happy thanksgiving.

 

[fastamdman]

1. I have never killed a chip before and that has nothing to do with not pushing things hard enough. I and my friends both are huge overclockers. Think liquid nitrogen, DICE, sub zero air temps, etc 😉.

As to your number 4 question though. Were you overclocking le hardcore while folding 24/7, that is what I do right now. I wouldn't say hardcore but my 2500k has been running anywhere from 4.5ghz to 5ghz with an array of voltages to "get things done" and there have been no issues, nor will there be. Chips don't just randomly stop working generally but they will encounter degradation. Meaning you need more voltage to do the same exact thing.

I did a lot of socket 478 overclocking back in the day, pushing the 2500m if I recall correctly was the "thing" to do. Stuck mainly with air back then and saw huge voltages but never killed a chip.

GENERALLY when I think of chips dying my first thought goes to condensation. You will see a lot of real overclockers use things to stop condensation from happening, but time to time every now and then something fails and you burn up a chip or a motherboard. Generally motherboards though, sometimes both.

Chips do die, but from the sounds of your original post you are talking more about degradation then a dead chip. Dead chips don't boot up, degraded ones do 😉

 

[soccerballtux]

1. I have never killed a chip before and that has nothing to do with not pushing things hard enough. I and my friends both are huge overclockers. Think liquid nitrogen, DICE, sub zero air temps, etc 😉.

As to your number 4 question though. Were you overclocking le hardcore while folding 24/7, that is what I do right now. I wouldn't say hardcore but my 2500k has been running anywhere from 4.5ghz to 5ghz with an array of voltages to "get things done" and there have been no issues, nor will there be. Chips don't just randomly stop working generally but they will encounter degradation. Meaning you need more voltage to do the same exact thing.

I did a lot of socket 478 overclocking back in the day, pushing the 2500m if I recall correctly was the "thing" to do. Stuck mainly with air back then and saw huge voltages but never killed a chip.

GENERALLY when I think of chips dying my first thought goes to condensation. You will see a lot of real overclockers use things to stop condensation from happening, but time to time every now and then something fails and you burn up a chip or a motherboard. Generally motherboards though, sometimes both.

Chips do die, but from the sounds of your original post you are talking more about degradation then a dead chip. Dead chips don't boot up, degraded ones do 😉

hm long have you been folding at 4.5ghz?

personally I would consider a chip that experiences prime errors a "dead" chip, IE you killed it, it's no longer suitable for use in a daily rig if it's giving you BSODs.

 

[nanaki333]

i have. it's not a long drawn out tale like yours though. it was probably the 3rd computer i ever built (back in 00), so naturally i was a know it all. put everything together and powered it up, but no POST. couldn't understand what was going on. powered off, fiddled with things, back on, nothing. i decided to let it run and maybe it just takes awhile with this mobo. yeah.... i started smelling burnt plastic... powered it off, took the HSF off and i left the stupid plastic thing on the bottom of the HSF! yeah.... it didn't work after that...

EDIT: it was a duron 700 i believe, or maybe 750.

 

[Idontcare]

I killed my QX6700 after running 24x7 loads on phase for 2yrs. It was running 4GHz ~ 1.55V. There was minimal ice buildup, I had it pretty well insulated.

But one day I decided to take it off of phase and use it on air (Tuniq 120). Reduced the OC to 3.3GHz, temps were low because I wanted less noise. But then one day it just died, refused to post.

At first I thought it was the mobo but I tried it in another known-good mobo and it still wouldn't boot (not even a post screen). Then I tried a known-good CPU in the first mobo and that system booted.

Dead CPU. I wouldn't have thought it would die on air at those clocks so I can only attribute its death to the extended use on vaporphase.

 

[fastamdman]

hm long have you been folding at 4.5ghz?

personally I would consider a chip that experiences prime errors a "dead" chip, IE you killed it, it's no longer suitable for use in a daily rig if it's giving you BSODs.

I have been folding at anywhere from 4.5 to 5ghz since I bought the chip new. A chip is not dead just because it gives an error in p95 or linx etc. It simply just means it needs more voltage, or the o/c is no longer stable at that given frequency. If we went by your standards ALL chips would be bad from day 1 because there is always a limit. Some SB chips can do 4.5ghz at 1.3 volts while others need 1.4 volts etc. It just depends on the chip. While testing though everyone gets errors. Once stability is found, errors go away 🙂

I killed my QX6700 after running 24x7 loads on phase for 2yrs. It was running 4GHz ~ 1.55V. There was minimal ice buildup, I had it pretty well insulated.

But one day I decided to take it off of phase and use it on air (Tuniq 120). Reduced the OC to 3.3GHz, temps were low because I wanted less noise. But then one day it just died, refused to post.

At first I thought it was the mobo but I tried it in another known-good mobo and it still wouldn't boot (not even a post screen). Then I tried a known-good CPU in the first mobo and that system booted.

Dead CPU. I wouldn't have thought it would die on air at those clocks so I can only attribute its death to the extended use on vaporphase.

If it's not broke don't fix it applies here heh. I would venture to take a guess that if you left it on phase it would have kept on trucking a bit longer. In my opinion it's kind of like a car. It might be hard to run it 24/7 at 4,000 rpm's, but it's also harder on it for the load to vary and for the engine to go from idle to full load at full rpm and back and forth etc.

 

[Dribble]

Never had a cpu die/degrade, however I've had motherboards, memory and psu's slowly go wrong.

 

[Rhoxed]

Never a cpu thankfully. Thought I had a Phenom 9550 die on me once, but turned out to be the MOBO (designed for 95W chips, and the 9550 was at 2.9ghz 1.475v for 3 years)

Got a cheap mobo for it, and it continues to chug along at 2.9ghz for my brother.

for reference (don't see the need to go back before x2's)
3 x 3800x2's - 2.5 - 2.9ghz
Phenom 9550 - 2.9ghz
Phenom 9850BE - 3.3ghz
2x Athlon II 640 - 3.6 - 3.8ghz
Athlon II 450 - 3.6ghz
PhII 940BE - 3.9ghz
PhII 720BE - 3.9ghz
PhII 1090T - 4.2ghz

 

[fastamdman]

Oh if bent pins counts as killing chips, I saw a person glue a heatsink onto the cpu and then when they pulled the heatsink off with a lot of force, it bent / broke some pins. Seen it happen when going cold too albeit we are generally smart enough not to use super glue as thermal paste lol.

 

[aigomorla]

dude im a cpu serial killer... :X

i need to stay hidden of i'll get busted.

 

[ArchAngel777]

I have never fried a chip. A testament to the quality of the Q6600 is that I was running it at 1.52v @ 3.6Ghz on water for two years, and then passed it on to my dad running it on air @ 3.33Ghz 1.42v for the last 2+ years... Total of 4+ years and is still going strong. Oh, and it was a B3. Additionally, I accidently forgot to plug in the fan at one point and ran it for a few minutes. Touched the heatsink and it burnt my hand. It was around 100celcius. I thought for sure I 'fried' my chip. Nope, she was just fine. Actually, to this day a Q6600 @ 3.33Ghz is still plenty of power for the games of today, even BF3. Impressive CPU in my opinion.

Been running my Core i7 920 @ 4.2Ghz for 2+ years now as well. Not sure on the vcore though, I don't remember.

I have been overclocking since the AMD K6-2 Days and the Celeron 300A... 😀

 

[Vic Vega]

The one and only CPU I fried was an Intel 486DX4 100Mhz and that was probably 15 years ago. I replaced it with an AMD Am5x86 150MHz which was much faster. 🙂

I didn't have the AMD long though and I moved on up to the Pentium. Oooohhh Ahhhhh.

 

[Cerb]

1. Have you killed CPU before?

Yes. After tiredly getting my Duron into a new mobo (should have kept the old one, because their deaths turned out to be systemic, and MSI made good on them), and went to get a bit to eat while the OS installed. I was tired, by this point, and really had not been expecting to need a new mobo in the first place, see.

The CPU fan was not plugged in.

2. If you answered yes to #1, please post the CPU architecture, frequency(s), voltage(s) that you killed it at

1GHz at a pretty minor Vcore bump (around stock Vcore for the same speed Athlons, but I don't recall it exactly...1.7-1.8?).

3. Give your best guess as to how you killed it.

No guessing needed. As soon as I looked inside, after coming back to a black screen, I did a pretty good self-hating sigh.

4. Were you surprised when it died?

Well, yeah, I was sort of planning on using it.

Surprisingly, I managed to have a rock solid Cyrix PR150+GP, which seemed odd for the time, and initially POSTed my above-mentioned Duron without a HSF on it.

 

[Zap]

1. Have you killed CPU before?
2. If you answered yes to #1, please post the CPU architecture, frequency(s), voltage(s) that you killed it at
3. Give your best guess as to how you killed it.
4. Were you surprised when it died?

1) Yes.
2) Intel 486SX-25, don't remember voltage, somewhere around 40-50MHz.
3) Trying to double MHz with extra voltage.
4) Nope. Killing it was half the goal (other half was 50MHz).

 

[Idontcare]

If it's not broke don't fix it applies here heh. I would venture to take a guess that if you left it on phase it would have kept on trucking a bit longer. In my opinion it's kind of like a car. It might be hard to run it 24/7 at 4,000 rpm's, but it's also harder on it for the load to vary and for the engine to go from idle to full load at full rpm and back and forth etc.

Yeah, I really should have left well enough alone on that one. The chip did take a lot of abuse by my standards. But nothing like the abuse of crazy suicide clockers that juice their chips with 1.7-2V for a quick 1M SPi run D:

 

[Rifter]

I've had caps go bad on a motherboard that then proceeded to fry the CPU, does that count? My poor AMD duron died like that.

 

[TakeNoPrisoners]

I have not killed anything yet, this is the first rig I have been able to overclock.

Running 4.0Ghz at 1.45V.

 

[Gillbot]

TL;DR--
1. Have you killed CPU before? No? Why aren't you overclocking harder then??
2. If you answered yes to #1, please post the CPU architecture, frequency(s), voltage(s) that you killed it at
3. Give your best guess as to how you killed it.
4. Were you surprised when it died? Were you overclocking le hardcore while folding 24/7? Well less of a surprise then right? In contrast, BD321 was surprised because he was within AMD's own recommended max voltage (1.5v, he was at 1.45) and his chip died.

TL;DR version of TL;DR version:
did you kill a cpu why were you surprised how did you kill it would you do it again

thanks to all, and to all (especially our Native American friends 🙁🙁🙁) a happy thanksgiving.

1: Yes, many many but i'll describe one of my "cherry" chips
2: Intel C2D, E6750 ~600fsb for I believe just above 3.7GHz, 1.45v core all others set to "auto"
3: Abuse? lol!!!
4: Not at all, I'm really shocked it held out for as long as it did. I never folded or anything on it but I pushed it up with the voltage around 1.55 a few times trying to see what the top fsb was. I determined that I didn't get much if any gain over 1.45 so I left it there. It was my main system which saw freecell, internet, email and gaming. I went through a few sup com games and ran well, BSOD'd later checking email never to run again.

 

[hdfxst]

i built a phase change unit and killed 2 xp1600's getting the mounting right,the first didn't have enough pressure and burned up before it got to the bios screen(i had 2 and didn't rma).the second one i had way to much pressure and had it in the bios screen heating it up to flow the tim and around 45c i turned on the unit and at around -30c it shattered into 3 pieces.my last socket a cpu i had was a 1ghz morgan core duron and i got that to 1700 and that cpu still works

 

[blackened23]

Anyone remember the 486 days with the ZIF socket? And some of them didn't have a "notch" to prevent incorrect CPU insertion. Yeah when I was a kid, I had an AMD 486 chip that I put in backwards. On an asus PVI/I486sp3 motherboard. lol, I don't know why I remember that. Anyway, it did not end well.

 

[jhu]

1. No
2. N/A
3. N/A
4. N/A

 

[LoneNinja]

I have never killed a CPU, but I have killed a motherboard. Handled my Phenom 9850 @ 3.0Ghz fine for over a year, put better cooling on it to get 3.2Ghz and the board fried less than an hour into Prime95. It was a fairly nice 890GX board too, expected it to handle a .1V increase on Phenom voltage, reviews overclocked with higher volts.

I personally don't push much past stock voltage either, the increase in power consumption just isn't worth it. For example my X4 965 is running 3.8Ghz with 1.375V right now, still need to try a lower voltage. To bump it up to 4.0Ghz requires 1.475V and increases power consumption of my system by nearly 40W, when the whole computer consumes less than 200W that isn't worth it for 200mhz.

 

[catilley1092]

I know that when you're folding, the CPU/GPU prefers a steady environment, it will run hot for long periods fine, too much variation causes problems.

Because when I started folding with folding@home, the people over there helped me a lot. I thought when a setup was new, to complete a job, let it cool, start over. They told me that the steadier the temps, the better, constant temp switching places more stress on the parts.

I gave up on it, because I was only GPU folding, and felt that a notebook could not handle the 24/7 thermal requirements for folding. I tried the newer folding app, where it uses both the GPU & CPU (the last generation 2.66GHz i5 & NIVIDA GeForce 425M 1GB DDR3). Although I was putting up the points (avg 6,000 per day), the Task Manager was constantly pegged (both the CPU & RAM).

A notebook computer cannot withstand that forever, although I hung in there for 4 months, w/o rebooting once, I felt that I done the right thing.

One day, when I can afford to build a high power desktop with proper cooling, I'll take a shot at it again. The folks at Stanford, who makes the Folding@Home possible for us to participate in, are good people. They're fighting to save lives. I will participate again.

Cat

 

[bononos]

I know that when you're folding, the CPU/GPU prefers a steady environment, it will run hot for long periods fine, too much variation causes problems.

Because when I started folding with folding@home, the people over there helped me a lot. I thought when a setup was new, to complete a job, let it cool, start over. They told me that the steadier the temps, the better, constant temp switching places more stress on the parts.

......

About the steady environment, is it about running things at a constantly high workload and avoiding the auto-freq throttling features of cpus in workloads which changes very frequently eg. games(?) - which I would imagine be stressful if the cpu were running at 100% then back down, then 100% and then back down...

 

[Subyman]

I had a huge power surge due to a direct lightning strike on my house's electrical system before. Fried the mobo, but the CPU was still good 🙂 I've never OC'ed it "hard" enough to destroy a CPU. I enjoy overclocking, but I don't abuse my equipment. Perhaps when I get out of grad school and can afford another hobby I might do exotic cooling methods and push harder.