http://www.johnshim.com/a64/cachetab.jpgQuote:
Originally Posted by cadaveca
Printable View
http://www.johnshim.com/a64/cachetab.jpgQuote:
Originally Posted by cadaveca
What was your max Prime95 stable overclock on air? on water?Quote:
Originally Posted by shimmishim
Here are some screenies of my XP-90:Quote:
Originally Posted by Magnj
XP-90 side shot
Another shot
Top shot
:shrug: ...........either one, but I reached my limit on my other NewCastle, not because of high temps necessarily, but the on-die memcontroller teetered out early in the race. :(
my max prime stable is 2640 which is what i'm running right now.Quote:
Originally Posted by Ackbar
i'm benchable at 2.87 with 1.67 volts.
i've noticed though that this chip get fast really fast!
under my prommy, i can superpi at 3 ghz but not more than 2.75 prime stable.
I've got a XP-120, I didn't realize how awesome it was until I compared it with stock temps. I went from ~60 on load to ~45 on load!
What do you mean it's getting fast really fast? It's getting faster as you use it/burn it in?Quote:
Originally Posted by shimmishim
thanks for the pics. I'll save the 10$ and pick up a panflow. Wasn't sure how far they hung over ram, looks godo to me. Thanks again. We'll have similar rigs minus the RAM and Video. Coolio yo
Oh go please tell me these get faster with burn in. That would be sick nasty
what temp should i keep the chip under? This is my first 939 build so i have no idea.
So far at 1.456v ive been prime stable for 25min at 2605mhz. This is on the stock cooler with AS5 instead of the crap TIM they come with. Temps reached 55*C
Thanks muchly shimmishim. Now how about postingthe same from CBId? Just wondering if they have disable the cache the same way as the sempron 2600+.
That's OK under constant prime with stock hsf. Keep it under 60°. I have the same chip from zzf and am testing nowQuote:
Originally Posted by boostdfd3s
Quote:
Originally Posted by Originally Posted by boostdfd3s
too many freakin settings in the DFI bios! :eek: So far the 2605 is my best. My 2x512mb mushkin BH-5 errors out in memtest test 8 at 230fsb @ 3.2v. Gonna try more in teh morning
Interesting that so many people are limited at 2600 (myself included). I think we may all have something limiting our overclock that we don't know about...Quote:
Originally Posted by boostdfd3s
I feel your pain :(Quote:
Originally Posted by Ackbar
Mine seems to hit a brick wall too, regardless of voltage.
I have another one on the way though :)
LOL... well, I don't have that liberty since I already bought this CPU to replace my crappy Winchester. Well, what I noticed is that this CPU really hates the heat. Maybe that's why it's so responsive to phase/water. I noticed that going from the stock HSF to the XP-120 alone got me from 2500 to 2600 or so. It's possible that this is the CPU you want for water or phase, but not necessarily for air... :(Quote:
Originally Posted by redgoo
Ok, Booted at 260 x 10 on my MSI k8N Diamond stock everything -- vcore I think only 1.456 in BIOS. Ran 10 minutes and wamred up and rebooted. I guess this board undervolts. Going to give it 1.5V vcore and try again.
I"m there too. Actaully, I haven't been able to get it prime stable at 2.6ghz. I haven't tested it above 1.6v, though. I've got a good WC setup, too.Quote:
Originally Posted by redgoo
clawhammers and newcastles should be safe up to 1.7 under water. remember that claws were 1.5 to begin with, right? i think these being fx chips is a very good possiblitity...
I've taken mine to just under 1.8V on air so far! :eek:Quote:
Originally Posted by cadaveca
Unfortunately... no change in stability.
well, then maybe like newcastles of 754, 2.6 is a wall. I have an older 3400+ that will do 2.8 with 1.725, but the one i am using currently, cbazc, will not go over 2.6 at all. stable @ 2597 1.65, but anything higher just will not work. tried different cooling, options, you name it..no go past 2597. Then i see that this speed is preset in the uguru software as well...i'm bios flashing and trying again at all speeds, and have 7 bioses left to try. something is preventing it from going higher, and i don't think it's the chip. winnies liked tccd...maybe these chips don't.
Well, if you get past it, let us know. It's oddly coincidental that everyone is stuck at around 2600...Quote:
Originally Posted by cadaveca
you bet it's weird, and what prompted the new attempts. I thought it was just my chip, but seeing these results here, with better steppings, kinda throws a chink in the gears. I really want to say it's a chipset or LDT limitation, based on what i have seen, but when i tell others that this is thier issue, i'm apparantly nuts, or a noob.
I for one havent tried to take it past 2600, simply because im still learning the DFI board...quite a jump from the NF7-S ;) That and the fact im on the stock HSF, dont wanna put too many volts in her
Well, what several of us have found is that voltage basically doesn't matter once you get to a certain point. So maybe any cooling will get you to 2600, but passing it is a problem.Quote:
Originally Posted by boostdfd3s
but if my cbazc stepping newcastle can do 2.6, then a caazc or caa2c should be able to go higher, at least a bit, no?
new newcastles (rev 0.6) also feature LDT x5(1000mhzx2) instead of the x4(800mhzx2), so maybe this change is what makes the 2.6 wall appear, as that is the only difference between my 2 chips. New revisions don't always mean higher speeds, unfortunately.