What are the pros and cons of overclocking

[Jeff7181]

Too many opinions here and not enough facts... it is a fact that you can get more performance by overclocking WITHOUT reducing stability. For those of you who like to try and prove people wrong... I did not say you too can overclock your XP2500 to 2.3 Ghz just like I did with no problems. Setting the speed higher is one thing, any idiot can do that... testing it for stability at that speed is another thing that requires partial use of your brain. For those of you handicapped in that area, don't overclock, you'll probably ruin your hardware or end up with an unstable computer that you'll blame on which ever one of the manufacturers that's it's "cool" to rip into at the moment.

 

[smashp]

When you overclock, you trade stability and reliability for speed and cost effetiveness.


the other thing overclocking does is gives you something phalic like to Brag about.


Plus if you overclock, you spend all your time just doing that: overclocking. ot only does it show your poor and a cheap a$$ but it also shows you are a loser with too much time on your hands.

If you actually need to use a computer for a real productivity function, dont waste your time. Buy the top of the line.

 

[Jeff7181]

You always have something productive to say don't you, Steve?

Oh, BTW... your first statement is 100% false.

 

[sciencewhiz]

Jeff, you second statement is 100% false.

You DO trade reliablility, even if you don't trade stability.

 

[ROcHE]

Originally posted by: Johnbear007
Barton 2500+ = 85$
Barton 3200+ = 490$

I paid 85$ and got a 500$ processor. Thats a pro to me!


On top of that it's fun! I have never had a component burn out on me yet.

Same here 😎

 

[magomago]

Originally posted by: sciencewhiz
Jeff, you second statement is 100% false.

You DO trade reliablility, even if you don't trade stability.

I think it depends...When I overclock I run prime95 for a week straight and memtest for 72 hours prior to overclocking to make sure all is well on the "stock" front...and then I overclock and after I got my overclock I do the same thing...

...and if its just as reliable....

🙂

And if reliability if losing an extra 2 years off its 10 year lifespan...so be it - it'll be obselete by the time I'm into something else

 

[fredtam]

Originally posted by: xerosleep
You can run into problems. I don't bother with it anymore. My old athlon would get to hot and lock up (didn't even overclock that much). Usually the system isn't as stable. My new intel won't let me overclock because it gets unstable. My brother overclocked his 2.8 to only 3.0 and it messed up windows XP until it was stuck in a continues reboot (didn't help to set everything to default again because it jumbled up the boot files). His friend had the exact same thing happen to his too. It's not worth it in my opinion.

Had it happen too ( corrupt windows). The difference is I knew what to do . Threw the XP cd in -repair-done. He most likely allowed the pci bus to get to far out of spec (should have locked it). Mine happened OCing the vid card ( wasn't paying attention to what I set the core clock to). If you don't know what your doing to begin with or don't know what to do when things go wrong then no you should not OC.

 

[dullard]

Originally posted by: magomago

I think it depends...When I overclock I run prime95 for a week straight and memtest for 72 hours prior to overclocking to make sure all is well on the "stock" front...and then I overclock and after I got my overclock I do the same thing...

There are many causes of a crash. Overheating is just one of them. Yes with overclocking you have to check if it overheats. But honestly I've never seen a computer crash from overheating (except when overclocking too much once). In reality almost all of the crashes are when opening/closing programs. So opening prime95 and leaving it running is one of the worst possible tests for stability. Try opening and closing all of your programs repeatedly for 72 hours and tell me if it is stable. And yes overclocking can affect that as well.

 

[Jeff7181]

Originally posted by: sciencewhiz
Jeff, you second statement is 100% false.

You DO trade reliablility, even if you don't trade stability.

I don't... maybe YOU do... but I don't. My computer is just as reliable as a computer with everything running at manufacturer's specs. Again, it goes back to overclocking properly... if you do it properly, there are no consequences except increase performance, increased heat, and decreased lifespan. The decreased lifespan isn't important to me because I definately don't plan on using this processor 3-5 years from now (I guess 3-5 years is about the normal upgrade cycle for the majority of people who aren't computer enthusiasts.) I actually don't plan on using this processor 6 months from now, so the decreased lifespan becomes even less significant.

*EDIT* If one wanted to get anal about it, one could say that if you increase the speed 1 Mhz, it's still overclocking... now do you think that would effect the stability and/or reliability of a 1,826 Mhz CPU?

So, I guess I should explain what I mean by "overclocking properly." When I overclock, I don't sacrifice stability, and I don't create an environment that will reduce reliability. Meaning... I don't feed it 2 volts... and I don't run it at 80 degrees C.

So... my XP2500 at 2.2 Ghz on the default voltage is running within specs that effect the lifespan of a CPU. Temps are within range, voltage is as AMD set it. Now I actually have it running at 2.3 Ghz on 1.700 volts (haven't updated profile and benchies yet) ... so the voltage is a bit out of spec... but the temps are still within spec. However. One could argue that even 1.700 volts is within spec, as some motherboards tend to overvolt, like the one I got for my aunt... default voltage is 1.650... the Epox motherboard was giving it 1.685-1.71... so you have to consider that AMD has probably allowed for these variances in voltages between motherboards... so mine, being set at 1.700 and varies between 1.690 and 1.715... isn't running THAT far out of what AMD intended.

 

[sciencewhiz]

Originally posted by: Jeff7181
if you do it properly, there are no consequences except increase performance, increased heat, and decreased lifespan. The decreased lifespan isn't important to me because I definately don't plan on using this processor 3-5 years from now (I guess 3-5 years is about the normal upgrade cycle for the majority of people who aren't computer enthusiasts.)

You admitted to a decreased life span. That is the definition of decreased reliability. You say that decreased lifespan isn't an issue for you, and that is true, but you can't say that because it doesn't apply to you, it doesn't apply in general.

If I tried to tell my boss that increasing the heat output of one of my devices didn't decrease the reliability, my credibility as an electrical engineer would go through the floor.

 

[fredtam]

Originally posted by: sciencewhiz

Originally posted by: Jeff7181
if you do it properly, there are no consequences except increase performance, increased heat, and decreased lifespan. The decreased lifespan isn't important to me because I definately don't plan on using this processor 3-5 years from now (I guess 3-5 years is about the normal upgrade cycle for the majority of people who aren't computer enthusiasts.)

You admitted to a decreased life span. That is the definition of decreased reliability. You say that decreased lifespan isn't an issue for you, and that is true, but you can't say that because it doesn't apply to you, it doesn't apply in general.

If I tried to tell my boss that increasing the heat output of one of my devices didn't decrease the reliability, my credibility as an electrical engineer would go through the floor.

Since the chips are rated to last 7/10/30 years or whatever at temps much higher than the average OC'er will reach (>70c ) it is safe to say that there is no decreased lifespan. Lifespan and reliability are two different things. Reliability means that it is going to work properly within its lifespan.

 

[Jeff7181]

Decreased reliability to me means it might fail tomorrow, it might fail next year... but nobody knows. I'll bet my whole computer that it won't fail tomorrow... or next week... or next month... or next year.