Have you ever damaged a CPU due to overclocking?

Have you ever damaged a CPU due to overclocking?


  • Total voters
    151

Bandalo

2[H]4U
Joined
Dec 15, 2010
Messages
2,660
Based on a discussion in another thread, I'm curious if anyone has actually killed their CPU from overclocking.

If you did kill one, did the manufacturer RMA it for you? How was the RMA process?
 
I did kill an Intel 2600K a while back, but it was out of warranty. The AIO cooler that I was using (an older Corsair H80, the original) had the pump crap out and the CPU must have got too hot and died. The CPU might have survived if I hadn't overclocked it.

I had two original H80s, an original H100, and a H115i all crap out on me due to pumps dying/gunking up. Only the H115i was under warranty when it died, the others were just outside of their warranty.
 
I degraded a Q6600 from 3600mhz to 3000mhz, took several years. Just about killed a Ryzen 1700x by over-volting - XMA on it no longer works. Other than that many cpu's OC successfully for a very long time.
 
I did kill an Intel 2600K a while back, but it was out of warranty. The AIO cooler that I was using (an older Corsair H80, the original) had the pump crap out and the CPU must have got too hot and died. The CPU might have survived if I hadn't overclocked it.

I had two original H80s, an original H100, and a H115i all crap out on me due to pumps dying/gunking up. Only the H115i was under warranty when it died, the others were just outside of their warranty.

Well, that's killed by lack of cooling though, not directly by overclocking. They probably would have died even at stock speeds under those conditions.
 
My only one was the 7700k I posted about recently, and I didn't really even do anything manually. I just ran AISuite, let it do what it did, and didn't bother to think about it. I don't know if I blame myself for that or not. The only ruining of expensive hardware I completely blame myself for was back in the early 2000s when I fried a video card because I didn't leak test my water cooling loop.
 
I killed a AMD K6-2 500mhz to 600mhz, it died like 5 seconds after post. That's the only CPU I've killed.
 
Well, closest I came to destroying was when I delidded a s939 Venice and there was as a result a tiny gap between the CPU die and heatsink base. System fired up twice, and automatically shutdown after 4 seconds.

Other than that I seriously damaged the VRM on my current motherboard (P35-S3G) because I stuck a C1 Q9550 in there and OCd it the f**k up. Throttlestop showed evidence of the motherboard skipping cycles in order to save itself but I decided to override that. Curently my VRM area is brown and warped :D works fine though!
 
I've killed various CPUs from various mistakes, but only saw one sort of 'die' from overclocking.

I had my 2600K at 5.2ghz with 1.52v under water for ages, but it eventually slowly started to get worse. In the end I was at 1.52v at just 5ghz truly stably after about 2-3 years roughly.

And I can't be 100% sure it wasn't the motherboard (that eventually died from pin breakage).
 
I can only remember doing it once. I degraded a Core i7 6950X tthat would run at 4.3GHz down to about 4.1GHz. Video cards on the other hand, are an entirely different matter. I've killed many of those.
 
I've killed a few Socket 7 processors and a Athlon Slot A 650. The Socket 7 CPU's were Intel 75mhz, 133mhz, 233mhz. Cyrix 166+. AMD K6-2 380mhz, 500mhz, and K6-3+ 450mhz.

With the Slot A I had "decovered" the processor to have full access to the processor with a Gold Finger device. As I was ramping it up, setting by setting to see how far it would push - well on the last GFD switch throw and startup it physically blew a small resistor off the PCB and hit me in the face.
 
As I said, I have only damaged one processor due to overclocking. Over the years I've worked with thousands of CPUs in one form or another and in that time I've only seen about 5 or so dead Intel CPUs and somewhere between 7 and 10 dead AMD CPUs over the last 20 years. These include all forms of damage from power surges, overclocking, or actual defects. Processors are some of the most reliable and robust computer components so long as they have adequate thermal protection and cooling.
 
AMD Thunderbird .... dead as a doornail ....

I had one that burned up the moment I fired it up. The CPU was brand new out of the box and I installed it correctly. I had good heat sink contact and the thing just fried on me. I got AMD to replace it under warranty. I still have that CPU to this day. It's an Athlon Thunderbird 1.33GHz chip.
 
Yep.. only 1 though.. Was an old Pentium 3 450 Mhz.. Tried to get it to 600 Mhz and kept upping the voltage.. started it up and about 2 seconds later got magic blue smoke and the distinct smell of an electronic components death.. Prolly overclocked 3-4 dozen computer builds since then and havent not had a problem since..
 
Older CPUs were a lot easer to kill, as they didn't have any thermal or power management.

Today, about the only way you can kill a CPU is going crazy with overvolting.

I've killed more CPUs from damaging pins or improper heatsink mounting (those fragile Durons) than I have OCing, but I've managed to pop a couple of the older units with OCs - probably more to do with long term thermal/cooling than the actual overclock.
 
Killed a 286 8MHz chip after soldering in a 16 MHz clock crystal, but I don't really count that one. Burnt a Pentium MMX, I think it was a 120 or something, by overclocking. It lasted about 4 months and then just died. I did not RMA the unit, because of other external circumstances at the time. I still have that MMX CPU on my desk right now, to make sure I remember....
 
Back in my Vapochill days, I killed an Athlon XP. Although, not exactly from an overclock, it died because I forgot to put grease in the socket. It is amazing how long they run before condensation fries everything!
 
I had one that burned up the moment I fired it up. The CPU was brand new out of the box and I installed it correctly. I had good heat sink contact and the thing just fried on me. I got AMD to replace it under warranty. I still have that CPU to this day. It's an Athlon Thunderbird 1.33GHz chip.
Maybe you cracked the core when installing the HSF??? I know *I* did that - not once but twice. Those were sad panda days...
 
Back in my Vapochill days, I killed an Athlon XP. Although, not exactly from an overclock, it died because I forgot to put grease in the socket. It is amazing how long they run before condensation fries everything!
I actually wrecked an Opteron 165 BY putting grease in the socket with my Vapochill LS. Bent the pins :(
 
Q6600 2.4 @ 3.2GHz used for years on end, then reverted to stock speeds. After enough years it would boot but wouldn't make it into an OS without extra volts even at stock frequency. Also cooked my poorly clocking Athlon 64 3000+, it was a terrible overclocker barely did 10% more than standard and also suffered degradation to the point it was useless in a small number of years/months.

I had an athlon xp-m also that did 2.8GHz plus at 1.9 vcore, I used it at 2.5GHz at low voltage because I didn't want to cook a good clocking chip.
 
VRM components on my Asus P8P67 Pro failed and killed my 2600k. One year after the warranty expired on the CPU. Not sure if OCing played a role. It was only at 4.4Ghz at the time.
 
i've yet to kill a processor unintentionally but i've killed an athlon xp 1700+, sempron 3000+ and athlon 64 3200+(clawhammer) suicide overclocking them for shits and giggles. i'm honestly amazed my Phenom II 940 didn't die when the VRM's on the motherboard i was using at the time shit the bed while it was running at 3.9Ghz.. these days though it's pretty hard to kill a cpu while overclocking.
 
I killed a 5960x by accidentally feeding it too much voltage, bought it used from [H] user a couple weeks prior. RMA'd it with intel and they sent me a sealed one and I sold it on eBay for more than I originally paid for the old 5960x.

Bought a 6950x used on ebay without the box and pushed it to high OC and it died, not sure if it was due to OC or my board killing CPU's. RMA'd with intel and they sent me a sealed one. Got out of overclocking at that point.
 
Right before they were released, I was attempting to push a Phenom processor on a GIGABYTE motherboard. I threw a fuck ton of voltage (1.45v I think, maybe more?) at a Phenom X4 9600 processor just to get a paltry 2.6GHz stable. Normally, we get some briefing on a processor, including what they are likely to do while overclocking. I had ZERO information from AMD on what was safe or not. I had no idea where the overclock ceiling was, nor what voltages ranges were safe.

Obviously, I didn't see a whole lot in the way of clock speed increases out of it so I started pushing the clocks and the voltages with almost reckless abandon. While overclocking, the VRM's on that GIGABYTE board shit the bed on me. About half the phases caught fire and there was smoke everywhere. The board actually generated an audible alert tone I've never heard before or since. I wasn't even aware that there was a mechanism in place to do that at the time. The smoke was so bad my office stunk for two days. As it turned out, the CPU was fine and worked on other boards without issue. (It still didn't clock for shit though.)
 
Never killed a CPU due to OC'ing but I did experience severe CPU degradation due to OC'ing a Nehalem i7 920 years ago.
 
Killed an Athlon but it was the cooler not installed properly rather than overclocking.

Had a 3770k degrade.

That's about it, generally speaking I don't put much stock in the arbitrary voltage limits people make up. If you've been round the community as long as I have you see that someone (respected) generally just picks a figure and then that's the accepted voltage cap. These days intel give More information. For me, as long as it's rock solid and cool I go with it. If it degrades after a long period of time then I'd just replace it.

Interestingly my i7 920 still overclocks like a beast, at the same voltage like 7-8 years later.

All that happened with my 3770k was that it needed more volts and I couldn't keep it cool enough. Delidded and i could cool it again and got the MHz back.
 
The only CPUs I've ever damaged in my life were a couple of 800MHz Durons back in the day when reviewing Socket A HSFs.
F*cking things would burn (improper contact) or crack if you looked at them wrong :|

Nothing beats the sickening feeling your stomach when you mount a Golden Orb and hear that distinct *CRUNCH* :(
 
...That's about it, generally speaking I don't put much stock in the arbitrary voltage limits people make up. If you've been round the community as long as I have you see that someone (respected) generally just picks a figure and then that's the accepted voltage cap. These days intel give More information. For me, as long as it's rock solid and cool I go with it. If it degrades after a long period of time then I'd just replace it.

So damn true!

I just go by temperature and stability. My last 3 cpus should have died in a matter of days if internet wisdom is to be believed.
 
The only CPUs I've ever damaged in my life were a couple of 800MHz Durons back in the day when reviewing Socket A HSFs.
F*cking things would burn (improper contact) or crack if you looked at them wrong :|

Nothing beats the sickening feeling your stomach when you mount a Golden Orb and hear that distinct *CRUNCH* :(

Or the turbine core one that you latched on... That one was always terrifying to put onto athlons....
 
Never from overclocking. I did start a PC I was putting together with my brothers Thunderbird 1.4Ghz chip in it with no HSF back in the day. On for maybe 5 seconds before I realized and pulled the power, but it was too late. Magic smoke escaped from the core.

Which always reminds me of this gem:

 
i think i killed my fx-8120. it seemed to be fine @4.8, until my mobo vrm fried, now it doesn't seem to work either...
so I'm not sure if i killed it or the mobo killed it.
 
Oh yeah, killed a couple tbird athlons myself, one b/c the fan that came on the stock HSF went out like a day after I put it all together. New retail boxed CPU. Theres no mistaking the smell from flaming silicon, that's for sure.

Kinda killed a P3 733 chip from overclocking. The L2 cache like stopped working. I verified that somehow but forgot by now. Everything including the OS was slow. Replacing the CPU solved this.

Have had a few Core CPUs actually degrade to the point that I had to up volts on stock speeds to maintain default clocks.
 
In the early days there wasnt an internet and overclocking was new. All the mistakes were my own.
Not really mistakes, just consequences of trying a bit harder and not necessarily voltage.
Some kit was bleedin flaky.

When you are the guy to ask you have to kill your own kit to know the boundaries.
 
I did kill an Intel 2600K a while back, but it was out of warranty. The AIO cooler that I was using (an older Corsair H80, the original) had the pump crap out and the CPU must have got too hot and died. The CPU might have survived if I hadn't overclocked it.

I had two original H80s, an original H100, and a H115i all crap out on me due to pumps dying/gunking up. Only the H115i was under warranty when it died, the others were just outside of their warranty.

With the Corsair Link software, you can create a rule that if the pump rpm drops to 0, then shut down the machine ASAP.
 
Nope never killed one during overclocking, they are pretty durable. I've actually only ever had a few chips die, usually due to pins falling off or purposeful excessive force. I think I had two that died from other causes.

One died from a foxconn socket issue with the second gen i7's

and the other my dog ate a while back (i think it was an i5.., but i don't remember)
 
Back
Top