• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

1.2 GHZ overclock! Any thoughts?

I finally finished building my rig about a week ago and decided to start overclocking. I ended up hitting 3.02 ghz with a x9 multi at 1.65v, 3x336fsb, and 216mhz for the ram at 1T. I'm very pleased with the results so far. Btw, it's running at these speeds 100 percent stable according to a bunch of sandra tests and I did not bother going past 3.02ghz.
 
1.65V is rather high, but then again who cares if you burn it out in a year or two. In the mean time you're laughing.
 
Originally posted by: chris21adan
It's a lil high but it doesnt reach temps higher than 48c so should I not be be fine?

temps are good.

try a program like prime95 small fft or superpi 32M.

not sure how good the sandra program is at testing stability.

and 1.65 is a little high... i try not to go above 1.6 on water/air.
 
No, electron migration is a problem with current gen CPUs aparently. You're killing it slowly by running it overvolted (temps would do the same thing but they are independant of each other, it's a different means of failure), but we don't know just how much of lifespan you're cutting from it by running at those voltages. There's a much better explanation than that on a thread around here somewhere...
 
Voltage is a much greater problem than temperatures. I'd rather have a CPU running at 55C than a CPU running at a 25% overvolt. Have you tried staying at 2.9 or 2.8GHz? Perhaps the voltage requirements for that are much lower.
 
I would stay under 1.5v but that's just me. One thing is for sure, though, 170MHz for .15v is not worth it.
 
Originally posted by: chris21adan
I can hit 2.85 with 1.48-1.5v. Should I just try and find a clock speed that enables me to not pass 1.6v?

That is the sweet spot you should be looking for...I consider 1.5 max regardless of cooling, and other then some synthetic tests I don't think you would notice the 200Mhz. I also doubt that you could run Prime at that speed/voltage and be stable...but I have been wrong before.

Anyway, congrats on a nice chip 🙂
 
I would have to agree. 1.65 is pretty high. You could probably easily keep it at 2.7-2.8ghz with ALOT less volts and not be killing the chip. You know how small performance difference 2.7-2.8ghz is compared to 3ghz? Its not even noticeable and you wouldnt have to kill the chip to get it. Thats just my opinion. I ran my opty 144 @ 2.6ghz with 1.43 volts. Plenty of speed for me.


Originally posted by: Furen
I would stay under 1.5v but that's just me. One thing is for sure, though, 170MHz for .15v is not worth it.

 
I think the thread Bobthelost is refering to is this one http://forums.anandtech.com/messageview.aspx?catid=28&threadid=1773169 where some of the Intel boffins (pm & dmens) describe the effects of over volting in some detail. Interestingly the biggest damage these days is not from electromigration, but from dialectric breakdown, among other things. The most important message to take away from that thread is that there is an exponential reduction in cpu operating life with increases in voltage, but only a linear reduction with increasing MHz and heat. High voltage will still damage your cpu even if temperature is within normal limits. However, if you only keep a cpu for a year or two, you can afford to take a few risks and push the volts up. I still think that 1.65v is a bit high although that is not based on any hard evidence. My personal limit is 1.55v which is just over 10% of the recommended operating limit for the Opty. Going to 1.65v for me only got another 80 MHz so wasn't worth it. See what you can get at your 'comfortable' voltage and go with that.
 
Good move. That's a great overclock for such a small voltage increase. I'm well jealous. What's the batch code, manufacture week for that chip?
 
Want to make sure this puppy's really stable? Prime95 for 9 or more hours ( I do it for 12 myself) and if you get any errors you're not stable.
 
Originally posted by: atybimf
Want to make sure this puppy's really stable? Prime95 for 9 or more hours ( I do it for 12 myself) and if you get any errors you're not stable.

qft

It's not stable if you don't prime test it.
 
Back
Top