Printable View
heh jamex, i just saw the discount on newegg as well, pondering if i should get one, but I dont have the money, if only i could sell my a8n32-sli for about 90-100 id be good, but as a student on a budget I can't get it :( even though its such a good deal!
Guess I'm sticking to my asus.
Put it on ebay man! If you are luck enough you'll sell it quick. Man, 50 bucks off.... I wouldn't let this one go away. I hope you can find somebody to sell the asus one. I'll let you know if the overclocks improve. Mine is gonna be here on wed/thursday, don't know yet, newegg is slow updating the tracking info. :mad:Quote:
Originally Posted by acrobuddy
i update the bios, and understant why i can't reach fsb 300 with my opty, my ram was low volted 2.55V and it's recommanded to used is @ 2.7V.
now i think i'm stable @ 300*9 ... more 33mhz htt and i hit 3ghz :slobber:
Jamex-
I hope the DFI gets you stable over 3.0GHz. Most motherboards change vcore when a load such as Orthos is placed on the system. My 32 board drops from 1.47 to 1.46v intemittently. May rise to 1.48v occasionally but it i fairly predictable. IMHO that is pretty good. You just have to know if the board dips or rises under load and factor that into whatever overclock you dial in.
I have a A8N SLI Deluxe that I stuck my 4400X2 (89w) in and the Vcore actually rises under load with that board. The problem is it will drop back down to at or below the dialed in Vcore. I believe that this is not a better situation because the Vcore rises way up above what the CPU needs in order to run.
As an example if the chip runs solid at 2615MHz 1.47v I have to take into account that there will be a rise to 1.52v. Not a problem till you approach 1.56v as the solid voltage for say a 2750Mhz overclock. Now the rise will take it over 1.61v which isn't good. It rides there for awhile but will then dip to 1.56v which is the lowest Vcore I can use for the overclock. Can't lower it to 1.51v so it will rise to 1.56v because it ineveitably dips to 1.51v (dialed in Vcore) and the test fails.
On the A8N32, for my 3006MHz overclock I set the Vcore at 1.48v and it dips slightly under load. Usually it is only .01v-.02v but I know what the ceiling is. Very short rises to 1.49v from 1.47v are seen but only for a second.
So, before you dial in your known voltages make sure you know what that board will do under load. Also, if you don't have a very good PSU you will see much larger variations in Vcore regardless of the mobo. Remember that an Opte at 3.0GHz 1.5v can pull some serious wattage, maybe around 200w I'd guess so a weak or stressed PSU will begin to put a lot of ripple and dips on the line which may be too much for any voltreg to handle.
Makes sense, but I think my variations are a little too big, going from 1.39-1.49 depending on load isn't good, I can rule out my PSU as I've tested with 2 others, my old Aspire 520W and my friends 700W GameXStream. Still does basically the same thing.
Woh!!! That is huge. Do you have CPU Overvolt enabled? IF overvolt is enabled and you are sure of your PSU (I'd test it with a meter, it's easy) then RMA the board. Something is wrong with the board if it has a .1v variation in Vcore.
My good friend and next-door neighbor as the same motherboard you guys go and his VCORE is all messed up (And this is his second A8N32)? Pretty much no matter what he does he can't get his VCORE voltage up above 1.45v (despite what the BIOS is set to)? If he enables the "overvoltage" in BIOS he can get a somewhat higher voltage to the CPU but then he has "warm-boot isssue" and can't reboot board!
Since I've never had this board I'm not sure what to do but I guess he should RMA again (even though first board had same problem)? Do either of you guys have similar issues to this?
Yeah I have it set to 1.25 with the +.2 over volt... hmm... seems my temps have risen a bit, isn't any hotter in here, in fact its gotten a little cooler here in Florida, but man I saw some nice temps when that last cold front came through, left window next to PC open all night.. haha I was seeing 28C idle!
ASUS told me to RMA, gave me a website to set up RMA, been 4 days and no contact after I set up the RMA, I don't want to call their number but I may have to.
Ok heres a steady idle vcore 1.456.. hmm seems normal, though I have seen it ~1.49 before.
http://img96.imageshack.us/img96/7390/1jx6.th.jpg
Ok load vcore, 1.408, hit 1.424 sometimes, but mostly 1.408
http://img391.imageshack.us/img391/1908/2ku2.th.jpg
Ok right after stopping Orthos, went back to stable 1.456
http://img294.imageshack.us/img294/9697/3oh4.th.jpg
And heres some goodies of airflow in my case:
http://img294.imageshack.us/img294/7...858ang3.th.jpg
That Tuniq 120 is huge!
http://img392.imageshack.us/img392/6...1859my1.th.jpg
Still worried about the temps there, I don't thing mount got messed, i may remount using electrical tape on the CPU for stability and protection and see what I can do.
That's no good mate. Can you stick a meter in a molex and watch the 12v voltages when you start Orthos? Red lead in a yellow wire hole and black lead in a black wire hole, set the meter to 20v DC and write down idle Voltage. Start Orthos and record the dip, if any, that you see. A $20 Radio Shack portable is plenty and comes in handy for a lot of things. Just trying to make sure it is actually the mobo before the pain of an RMA.
Quote:
Originally Posted by Merc14
Hey thanks man, I hope the DFI improves the overclocks, I have the feeling it will, I will let you know. That will prove they are "THE MOBO" for overclocking. If not, and is the same as the asus, I'll probably return it to newegg and keep the asus since it has the SLI X16.
Roger_D25-
The board goes up to 1.5625 in straight Vcore. CPU Overvolt adds .2v to that giving you a possible 1.725v. CPU Overvolt also stabilizes the Vcore greatly. I have no idea why but it is a must do if you overclock.
If your buddy is overclocking he should absolutely have CPU Overvolt enabled and just make sure to subtract .2v from the Vcore he sets. I suspect that he needs to flash to the 1303 BIOS. . It is a very stable version and drops the mobo temp by about 10c as well.
Jamex-
DFIs are definitely superb overclockers. The 32 is Asus trying to match them. In that they have done a good job, for the most part, in that there seems to be very little difference in the results but usually the top overclockers are running a DFI. They have their quirks, though, and are very particular about RAM and a few other things. Check out DFI STreet for more info.
Okay! Idle is 12.4V... load is 12.37V, but this is a separate rail then what my CPU is on, does that matter?
If I look, I see ASUS probe 2 is off on 12v by .24V, and let me see 5V.. 5.04 on multimeter.. so off by .12 ... hmm halved... maybe to check the 3.3V rail.. 3.32... so its off by .09 .. not so much a half... but i can guess that it also reports the Vcore wrong as well, reports the same as cpu-z btw. These are all idle tests.
Hmm tested the voltage off the CPU-1 connector, the 4pin 12v connector.. idle 12.41, load 12.34 .. not much difference.
12.41 to 12.34 is very solid. No worries there. The sensors on every board are not to be trusted, that is why I always suggest the meter check. If you want to be certain you can stick the probes in the 24 pin power connector but this is dangerous as a slip could short something.
The only other thing I could suggest is to see if the MOSFETS (voltage regualtors) are overheating. If cooling is poor around the CPU area the voltage regulators will overheat and you get very unstable Vcore. Once again, this is true for any board and the only way to tell, without expensive equipment, is to carefully touch the heatsinks covering them. They should be warm but not hot under load.
I had a bad time with the voltage regulators when I went to watercooling and lost the HSF fan in that area. I also removed the heatpipe since i have a pseudo BTX case and the 32 heatpipe doesn't work well upside down. Couldn't overclock more than 4%. Ended up installing a 120mm in the side panel after some cutting and sanding and stuck some heat sinks on the MOSFETS and all is well now.
You should be OK since you are on air but remember that you are stressing the voltregs and they need extra cooling under load as does any electrical circuit. So give them the touch test and if they are just warm then I guess you RMA the board.
Hope that helps. It sucks to RMA something and have the same problem because it was some other component screwing up.
I managed to overclock a cpu that aint on that list, Opty 165 Stepping: CCBWE 0551 WPMW. Managed to get it to 2.872GHz on air with a TTBT, not sure bout temp, think its about 50*C. Super Pi 1.5 29.328s 1m. 1.450v not prime stable (well not tried :P)
get it prime stable then you can be sure it's a complete overclock.Quote:
Originally Posted by Freebo
occt test is accepted ? :rolleyes:
use both if you're not sure. dedicated a few days to utilize prime and occt, personally i just use Prime, 3dmark2001,03,05,06 & some good ol Folding burns.Quote:
Originally Posted by madcho
Ok, im currently getting 2.6ghz primed, seems to be ok, not crashed yet, ill go onto the 2.87GHz soon, possibly at the weekend.Quote:
Originally Posted by likewhoa
awesome my premium lapping kit from easypckits.com just arrived today, 2 day usps shipping! anyway lapped my tuniq down to a very nice shine, much better than before, used some masking tape for support around my core, now the tuniq doesnt swivel! brought my temps back to the way they were 32c idle, but i lost 3c on load, load is now down to around 45c, used to be 48-49C.. just awesomeness :woot:
would have had pictures, but no camera :(
cant wait for this stuff to set, maybe i can hit 3ghz air? eh? eh? :stick: :slap:
What stepping is your 165? 1.45v for 2.9ghz sounds like my 0617. I'm also planning to get a tuniq.
hi!
i recieved my 165 step CCBBE 0615 EPMW. I had some problems with x3 LTD multiplier (cant go over 2,4GHz). Now with LTD x2 im testing 2,5GHz with stock cooler and 1,312v.
http://img474.imageshack.us/img474/2...rimezt7.th.jpg
Temps are a little high, but i want to try the max it can do with the stock, later i will try with my XP-90.
Hope to do 3GHz, lets see...
Updates soon :)
Acrobuddy-
That's looking great mate! Well done with the improved cooling. Lot's of work but 3c at load is a very nice buffer to have. What is the ambient air you are working in? My stepping pulled 2910MHz at 1.39v, 42c load in 74f ambient. The lower voltage may be due to the slightly cooler temps so you look to be right on course for 3.0GHz.
uF4No-
CCBBE is generally the good stepping and your early results look good for a 3.0 run. To make life easier you may want to download Orthos. It is Prime 95 with a nice GUI and runs dual threads of P95 by default. Much easier to use and control and you get the same results. http://sp2004.fre3.com/beta/beta2.htm
i have a CCBBE 0615 EPMW stepping running in ambient 77-80F, but I'd like to note that here in Florida, USA a cold front is about to sweep through, nice cold temps which will lower my ambient to around 70F! YES!