Does overvolting shorten lifespan if heat is under control?

S

smoothjk

New Member
Just as the subject/title line says...

Is overvolting above factory settings bad for the CPU if you upgrade to an aftermarket cooler that keeps the temps relatively low? Is there a harmful side effect other than just heat (i.e., electrical)?
 
Yes, the more stress you put on anything the less likely it is to last longer. That is just simple logic. Sure you can have it over clocked and super cooled for 5 years, but that doesn't mean you are running a lesser risk of hardware failure.
 
Hmm...do you think raising the stock voltage of 1.4 to 1.45 is conservative enough to still be safe and reliable (assuming, of course, that temps are under control)? I've heard of people going to 1.55+, so my goals seem reasonable, I think.
 
smoothjk said:
Hmm...do you think raising the stock voltage of 1.4 to 1.45 is conservative enough to still be safe and reliable (assuming, of course, that temps are under control)? I've heard of people going to 1.55+, so my goals seem reasonable, I think.
Click to expand...

It doesn't guarantee it will fail, you just run a higher risk is what I am saying. It may or may not fail, but when you over clock you just put yourself at a higher risk of hardware failure and stability issues. However, with higher end cooling it will help level the playing field, but nothing is guaranteed.
 
BUT the ODDS are, that the processor will continue to work well until it has been replaced or the computer is destroyed by another part.

lol tlarkin, i realize that some processors are more prone to break when overvolted, but why the 'i can neither confirm nor deny' stuff? :P
 
lovely? said:
BUT the ODDS are, that the processor will continue to work well until it has been replaced or the computer is destroyed by another part.

lol tlarkin, i realize that some processors are more prone to break when overvolted, but why the 'i can neither confirm nor deny' stuff? :P
Click to expand...

Since there is a margin of error that can be narrowed down to the user. I have no idea if this person can do everything necessary to make sure they over clock it in the right way.

Even then there is no guarantee with technology. I have had some over clocked things fail on me in the past, but then again over clocking technology and cooling technology has gotten a lot better so it is easier to over clock but that still doesn't mean you are running the risk of hardware failure.
 
the fact of the mater is, overclocking the piss out of your processor may cut its life in half.. but half of the procrssors normal life, is three times longer than you would EVER want to keep one processor. It will be so out of date by the time it goes bad. I wouldnt let it bother you.
 
Is overvolting above factory settings bad for the CPU if you upgrade to an aftermarket cooler that keeps the temps relatively low? Is there a harmful side effect other than just heat (i.e., electrical)?
Click to expand...
Yes. It's a phenomenon called electromigration that causes semiconductor devices at high voltages to die, no matter how low the temps are. Some parts are more likely to die because of this, some cope with higher voltages better. Generally, smaller manufacturing technlogies don't cope as good as previous ones (e.g. 45nm chip would fry pretty quick at stock volts of a 250nm CPU regardless of temps, and 65nm CPU can generally take higher volts than 45nm, though the specifics of a manufacturing process affect this as well)
 
Good point, but due to the relatively high life span of interconnects and the short product lifecycle of most consumer ICs, it is not practical to characterise a product's electromigration under real operating conditions.
 
Good point, but due to the relatively high life span of interconnects and the short product lifecycle of most consumer ICs, it is not practical to characterise a product's electromigration under real operating conditions.
Click to expand...
True, unless someone releases a new Northwood it probably won't matter unless you're a nut cooling with phase change or liquid nirogen... thenagain, if the OP was that particular type of nut, he probably wouldn't be asking in the first place.
 
Although overclocking has become a relatively safe procedure, you are still running a chip higher than the manufacturer set it as. Lower clocked CPUs that use the same chips as the faster CPUs either means that the lower clocked CPUs couldn't handle the stress of a higher clock for extended periods or they just lowered the clock of a batch in order to provide stock for a cheaper CPU. In either case the CPU was made to go faster and was set to a slower speed for a certain reason. This makes overclocking pretty safe since the CPUs theoretically could handle the speeds of the faster CPUs. Of course this still does put more stress on it which will make it more likely to get unstable, which is what will usually be the first thing to happen before total failure (i.e. broke). Cooling helps lessen the stress of heat but it still doesn't take away the fact that there is still stress from the high clock speeds. As an example, I had an old Athlon XP 2800+ and no matter how cool it was, it was unstable and gave errors at a certain overclocked speed. If you are willing to learn how to overclock correctly and figure you won't be using your CPU for a long long time, then overclocking is a great way to give you increased performance at little to no cost.


Good point, but due to the relatively high life span of interconnects and the short product lifecycle of most consumer ICs, it is not practical to characterise a product's electromigration under real operating conditions.
Click to expand...
True, but with some CPUs able to go above 4Ghz, electromigration could come into play there. Especially since with the smaller 45nm sizes, cooling is easier so people want to increase the voltages even more which only contributes to the problem.
 
Last edited:
Thanks for all of the helpful responses. I definitely don't plan on increasing the voltage by much, and I also don't plan on keeping my CPU for a terribly long time. I'm more concerned with my mobo or other components being affected.

Either way, I'm a conservative guy, so I'll play it safe.
 
If your cooling fails when you over clock it can result in catastrophic failure. That is what I mean when I say you always run a risk. Also, if you over look something you can screw it up yourself, hence the higher risk of failure.

The bottom line is, over clocking is a higher risk of hardware failure and not always a real world performance increase. Sure you may benchmark better but do you really notice a difference in surfing the web and typing in a word processor or playing music or movies? Even video games generally have little notice of performance increase when over clocking because games are designed by the programmers to run on the middle of the line systems.
 
tlarkin said:
If your cooling fails when you over clock it can result in catastrophic failure. That is what I mean when I say you always run a risk. Also, if you over look something you can screw it up yourself, hence the higher risk of failure.

The bottom line is, over clocking is a higher risk of hardware failure and not always a real world performance increase. Sure you may benchmark better but do you really notice a difference in surfing the web and typing in a word processor or playing music or movies? Even video games generally have little notice of performance increase when over clocking because games are designed by the programmers to run on the middle of the line systems.
Click to expand...

You could say the same thing about buying faster CPUs. You bench higher but you may not see a difference unless you do things that actually require a CPU to work hard. Although with a faster CPU, you don't have to overclock but it is basically the same thing.
 
gamerman4 said:
You could say the same thing about buying faster CPUs. You bench higher but you may not see a difference unless you do things that actually require a CPU to work hard. Although with a faster CPU, you don't have to overclock but it is basically the same thing.
Click to expand...

Yup and that is the exact reason I used to want to rip off the heads of sales people at my old job many years ago. I worked IT for a company that did service and retail/business sales. I worked on the service side and did warranty repair and builds, and OEM stuff for our customers. Some sales guy would sell some customer who just wanted email and web browsing a 1000 dollar computer with all kinds of features and upgrades. Then come to me and yell at me for it not working only for me to find out they just want a simple computer. 99% of the time it was due to them not knowing how to use a feature they didn't want in the first place....
 
tlarkin said:
Yup and that is the exact reason I used to want to rip off the heads of sales people at my old job many years ago. I worked IT for a company that did service and retail/business sales. I worked on the service side and did warranty repair and builds, and OEM stuff for our customers. Some sales guy would sell some customer who just wanted email and web browsing a 1000 dollar computer with all kinds of features and upgrades. Then come to me and yell at me for it not working only for me to find out they just want a simple computer. 99% of the time it was due to them not knowing how to use a feature they didn't want in the first place....
Click to expand...

lol. I hate people that try to get others to buy crap they don't need.
 
You must log in or register to reply here.
Back
Top