How much is too much

[b1lk1]

I say if it is a problem for someone, then just don't do it. I have been running my CPU around 1.65V for some time and I just make sure my temps are in check. If I wake up tomorrow and hit my power button and nothing happens, the I guess I should not have overclocked my system. I also set my 5V jumper on this board the day I got it and it will stay there too. My ram runs @ a constant 255MHz @ 3.44V and it will stay there until it too dies. If I was worried, I'd sell it all and buy a Intel MOBO and a P4. Just set things where you are comfortable. As long as temps are good, you are at least making sure not to hasten the death of your components.

[esco]

the age old rule comes to mind...

 

"if you cant afford to replace parts, dont overclock."

 

:angel: im in a lucky position and i can afford to (young, no house, no car, and full time job)

 

saying that if you cant afford to replace it but want to overclock then be more carefull lower your voltages and oc till you hit a wall. then try more voltage see if it helps then use your intuition as to wether it would be worth leaving extra volts in.

[Guest Lux]

So would you say that a 2.2Ghz 4400+ at 2.6Ghz @1.520v @38C full load is ok?

[esco]

in my opinion yes. maybe not according to others.

[naddie]

I'm currently requiring 1.6v to reach 2.62GHz on my x2 3800+. Too much?

 

My other question is the chipset voltage. No one asked about that. How much is too much? I had it set at 1.59 as a way to test if my unstable 3DMark05 runs are caused by not enough voltage. I need a guidline as to how far can i take it.

[Thasp]

PS I overclocked celery's for years using excessive voltage. And they still run flawlessly!

 

Probably because celery, like many other vegetables, does not conduct electricity.

 

Can you tell I hate "winny, sandy, vennie, celery"

 

Unless someone runs it 24/7 like that for a long time, we'll never know if it hurts the processor. By the time we find out, it'll be obselete, so who knows?

[Death909]

If you have to jump from 1.4v to 1.65v for just an extra 100mhz, I wouldn't consider it worth it. 2.4ghtz is fine, you won't see a huge difference between that and 2.5ghtz. I can barely tell the difference between 2.45ghtz and 2.85ghtz on my 146 while playing any games. For me, it takes 1.45v (the minimum I use) for 2.8ghtz, but for an extra .075v I can get 2.9ghtz. My RAM happens to be unstable around this point, but after 2.9ghtz it would take even more voltage just to hit 3ghtz. Same with my 3200+, 1.45v was fine for 2.6ghtz, 1.525v was fine for 2.7ghtz, but even with 1.7v, 2.78ghtz was not stable (temps were 40c to 43c at load).

 

There was an excellent guide somewhere on over-volting. One of the things it suggested was that you write down how far you go with 1.4v, then boost it to 1.425v, and so forth till you've gone as far as you feel safe. If you make a graph out of that, you will see that extra voltage gives good gains till a point, and at that point diminishing returns kick in.

[naddie]

Kinda disapointed that I can't keep my 2.62ghz at 1.6v

 

If i rem correctly, i think the most I could do at 1.4v was 2.2 or 2.3 ghz.

[JVG81]

First off, let me say this is one of the best posts I've read on here in a long time Lux. Thanks Now there are two factors in overvolting. One is that cpu's have much longer llife than we need them, so it's really ok. The other is that getting that extra 100Mhz on top with hefty amounts of volts isn't going to do much for system perfomance anyway.

 

How does temperature affect electromigration? you can run your CPU at 1.9v on phase change cause it never goes over 20C, er whatever, and you never hear from those guys on how long their processor lasted.

 

Higher temperature causes greater resistance, hence faster 'erosion' on the processor. This erosion is the same effect you get from raising voltage (running more current wearing off the tiny paths and lines on the cpu).

 

It would be interesting to see a sticky in these forums on how long past processors have lasted on increased voltages and what their voltages were. Like the O/C database.... thats great and all but are all those systems posted in there still running? Past systems that have failed and how long they lasted?

 

I would give my right leg for a sticky like this.

 

Well, sorta.

 

Just dan't see it happening because

 

1) Ongoing release of new cpu's (meaning new processes) makes a good comparison over longer periods of time difficult. Although the 90nm process is going to stay at least until after socket M2 on AMD's side so it should be worth it in this case.

 

2) Almost everybody who cares about voltage is too afraid to run the test. The enthusiasts with ultra high end cooling equipment who dont care as much about volts are no good study material either. They have fluctuating voltages (dont keep them at one level for benchmarking reasons or whatever) and a tendency to buy new cpu's quicker. Keeping track of what they do in a year's time is apparantly just too much otherwise they would have done it already for us mere mortals.

 

Sometimes you wonder if someone somewhere started this craze so that you could be exited for 6mo. and they could sell more boards/processors/ram/graphics cards and charge more for them. Just something to think about. I hate playing into anyones scheme.

 

What do you mean with exited for 6 months? You mean they dont want you to OC too much so the consumer will still buy some newer model?

 

I think there is definately some deeper thought into the release of new electronics. Just look at cell phone's for example, every month there is another one with slightly better specs - but I'm sure they can release a 5 MegaPixel 2Gb memory phone with a battery that still lasts a week and perhaps even a microwave or hot air oven if manufacturers wanted to today. It's just that they can't sell yesterdays phone anymore if they do this, and want to sell every little tech increase in a sepreate new phone if they can.

 

It would be nice if someone knowledgable enough to answer these questions subscribed to these forums. You'd think with the tech behind these boards and their rep this would be the place you'd find someone like that however most electrical engineers probably run their computers at "spec" because the little performance gain isn't worth all the other headaches... not to mention most of them are too busy with work to game or care about a 10-20 fps boost.

 

Weve had several people claiming to know techies at AMD and were gonna ask them something, but for some reason they never report back. Makes you wonder if if they are just full of crap or if voltage questions are one of the best kept secrets from processor bakers.

 

Speaking of that, when you can get a 10-20 fps boost from overclocking and your system already is running at 170fps.... what's the point?

 

Here is where the fun part comes in. But I think there is more to overclocking then gaining a few fps, it's also a bit of future proofing and getting the max out of your money - which is a good thing. Spending a lot to get this bit extra seems like an illogical thing to do suddenly.

 

Its like from the movie Fear & Loathing in Las Vegas.... "Once you get locked into a serious drug collection.... there's a tendency to push it as far as you can."

 

You should read the book too, it's .ing awesome. Makes you realise how good Johnny Depp and Benicio del Toro did in the movie.

[Guest Lux]

exited = excited sorry for the grammatical error, heh.

 

P.S.

<3 that movie.

[Fight Game]

IMO, 100mhz (or 4% increase in this case), is nothing you'll ever feel and not worth all the extra voltage. If you were going for highest overclock award, and needed to post some awesome screenshots of benchmarks, and didn't care if you fried it, then it may be worth it TO YOU. Maybe sell it for $50, and buy a venice 3000+ for a little more than $100 and goto 2600-2700 on air, easily. This question is really just a matter of opinion. You just have to weigh the advantages of the 100mhz (to you) vs the possibility of losing your cpu a little sooner, possibly very soon.

[Guest Lux]

My x2 is hitting 2.8 boots into windows and runs benches at 1.523v but i'd have to up it to make it prime stable.... great chip apparently... maybe not as good as some but still a good chip.