OC I7-3960x C2 & ASRock X79 Extreme6

[thebellbell]

Only if the voltage is increased. Voltage kills CPUs faster than heat does.

 

FRom multiple sources, I've found that 1.35 is being repeated over and over by ppl with the 3920k and 3960x, with an OC of 4.5 and what are your guys thought on using the EZ OC? Ill brb, while you think of that, I need to update my BIOS, its 4months old and the 2 newer versions that it upgrades to, upgrades stuff to do with USB and CPU Code, and I'm pretty sure my I7-3960x C2 is pretty new!

[Tjj226_Angel]

Only if the voltage is increased. Voltage kills CPUs faster than heat does.

 

I have read several articles that says that is a myth and has never been completely confirmed. I am not going to say you are wrong, but can you provide some evidence for my own knowledge.

 

I have actually tried to find charts of CPU lifespans and how it is affected by overclocking, but the closest thing I found is a guy with a q6600 running at 2.9ghz since launch day of the chip.

[ir_cow]

Fact: voltage kills Cpus i;ve seen it

Fact: heat kills cpu i've also seen it.

 

under normal settings I would say a cpu at say 1.4v 70c would die faster than a stock cpu at 1.3v @ 70c. Heat only kills if you are close to it's upper limit.

 

Was it bonker who just retired his i7 950 because it's been overclock so long that it's no longer stable? someone on here like last month had tread about it, duno who anymore.

Edited July 16, 2012 by hornybluecow

[Tjj226_Angel]

Fact: voltage kills Cpus i;ve seen it

Fact: heat kills cpu i've also seen it.

 

under normal settings I would say a cpu at say 1.4v 70c would die faster than a stock cpu at 1.3v @ 70c. Heat only kills if you are close to it's upper limit.

 

Was it bonker who just retired his i7 950 because it's been overclock so long that it's no longer stable? someone on here like last month had tread about it, duno who anymore.

 

It was probably due to something other than overclocking then. My i7 950 took a hell of a beating, and it is still going strong at 4ghz. Someone else now has it but from what I hear there hasn't been a single issue

 

If voltage kills, I want to know by how much. Is it a 1 to 5 percent life span loss, or is it more like a 10 to 20% loss?

[ir_cow]

I not willing to find out I like my working computer

[Tjj226_Angel]

I not willing to find out I like my working computer

 

Lol, I bet.

[Waco]

It was probably due to something other than overclocking then.

Of course it was due to overclocking.

 

The degradation happens because of both temperature and voltage (with the latter generally causing damage faster than the former). The chips are rated to run at their maximum temperatures for years on end without any issues, so any amount below that maximum temperature will likely extend the life and anything hotter will cause it to die more quickly.

 

Voltage will kill a CPU over time, this isn't new. Remember the "sudden Sandy Bridge death syndrome" from right after SB came out? People were pushing too much voltage through them and after running stable for a few months they'd just die out of nowhere. The newer chips tend to be a bit more resistant but voltage still causes damage over time.

 

 

You can't really place a number on it because it completely depends on what voltage you're running, how hot you're running, and how often you leave it loaded down so it's running at full volts and temps. That make sense?

[Tjj226_Angel]

Of course it was due to overclocking.

 

The degradation happens because of both temperature and voltage (with the latter generally causing damage faster than the former). The chips are rated to run at their maximum temperatures for years on end without any issues, so any amount below that maximum temperature will likely extend the life and anything hotter will cause it to die more quickly.

 

Voltage will kill a CPU over time, this isn't new. Remember the "sudden Sandy Bridge death syndrome" from right after SB came out? People were pushing too much voltage through them and after running stable for a few months they'd just die out of nowhere. The newer chips tend to be a bit more resistant but voltage still causes damage over time.

 

 

You can't really place a number on it because it completely depends on what voltage you're running, how hot you're running, and how often you leave it loaded down so it's running at full volts and temps. That make sense?

 

Yes, but I am just saying that in relation to many other i7 950s and other various 1366 socket CPUs, boinkers chip died rather quickly. There is a guy I talk to down town who has had his i7 920 overclocked while running about a 1.38 CPU voltage since a week after the launch day of the 920, and it is still going strong.

 

All I am saying is that for boinkers chip to die so early on in it's life when compared to several other chips that are overclocked (some of which are older and less stable) to me at least it seems like there are other problems than JUST overclocking that caused his chip to fail.

[wevsspot]

While I'm not generally a big fan of Anand - back in 2008 he did an excellent write up on overclocking. It was geared towards the 45nm processors, and specifically the E8500. I was initially interested in his write up because I had just bought an E8400 and this was one of the first articles that popped up when I did a search. It was a good read and I bookmarked it for legacy purposes.

 

Beginning on page five;

http://www.anandtech.com/show/2468/5

 

While the processor family is different than what's being discussed here, the explanation and risks are for all intents and purposes the same.

[El_Capitan]

I'm a firm believer in keeping within AMD or Intel's max voltage specifications. I also think temperature is a bigger factor in degradation. My i7-3930K is running 1.376V - 1.400V's since I've had it. My i7-2600K/2700K's have been known to go up to 1.48V's at 5.0GHz runs, but mostly 1.46V's at 4.9GHz. Never had any issues, yet. Max Load temps are always in the lower 70's to high 60's, while daily use max temps are usually in the high 50's and low 60's.

[wevsspot]

I agree with high temps being the worse of the two evils. As temps increase more, electromigration occurs at a greater frequency due to the electrons moving faster and that may/will eventually lead to burnt tracers on the silicon. So what affects temps?

 

Voltage

Frequency

Cooling solution

Thermal interface material

Ambient air temperature

Cleanliness of the system

Airflow inside the case

 

It's all a vicious cycle

[Waco]

I agree with high temps being the worse of the two evils. As temps increase more, electromigration occurs at a greater frequency due to the electrons moving faster and that may/will eventually lead to burnt tracers on the silicon. So what affects temps?

I'd bet I can kill a CPU with voltage a lot faster than you can kill one with heat.