Looking forward to the results...Anyone know where I can get a vapo or sumpin...?Quote:
Originally Posted by chew*
Printable View
Looking forward to the results...Anyone know where I can get a vapo or sumpin...?Quote:
Originally Posted by chew*
Sorry I couldn't make it bro...I still feel like ass. I meant to take it easy last night but the vodka tonics just kept comin... now I'm paying for it. :(
Best of luck today! Maybe in the near future Dumo, you and myself can get togethor.
Hate to keep you guys in suspense but I had set a goal for myself and I refused to not acchieve it......Persistance pays off in the end.
http://valid.canardpc.com:80/show_oc.php?id=492149
Results up on page 1 with a picture show in a little bit.
Running memtest to make sure I didn't lose a stick.
Nice job buddy! 5.0GHz ftw. :up:Quote:
Originally Posted by chew*
I finally started tweaking tonight, can't give away the results yet and I'm far from done, but let's just say our chips are more similar than you thought lol.
I started off the night finding max oc on the chip just by upping multiplier and voltage, and that was fun. Now I'm on to max northbridge clock. Max at stock voltage seems to be around ~2650MHz. Do you know if vcore helps nb clocking as well? First time ocing Phenom so I'm learning as I go. :)
I'll post my thread tomorrow night with my first results.
Nah you should have seperate voltages for NB, and a little tip when going for max cpu validation underclock your NB, lot of guys dont realize MHZ > NB and you lose max clocks when pushing the NB.Quote:
Originally Posted by EnJoY
Thx, I think thats the first 5gig I've seen on DI and on a cheap pot to boot.
Page 1 is completely updated. :up:
Nice going chew* :up:
It is first 5GHz I'm seeing on DI :clap:
So can you know play some 3DMark :p:
Yea, when I was going for max OC I used only the multiplier, so nb was always at stock. Temps only became the issue when I went to 1.65v which seemed to have no benefit whatsoever over 1.575v...same stability. Essentially 1.6v in bios is the max these chips will benefit from under >ambient cooling.Quote:
Originally Posted by chew*
Once I get my max nb figured out, I'll get some high cpu/high nb combination clocks and run some benchmarks. One thing I was amazed about so far in my early testing was my attainable HTT frequency. I was able to hit 290MHz HTT w/ HT/NB multi's both at 9x. This achieved me around 2700mhz which was at the edge of stability, but both seemed fine with it. I later raised the NB multiplier in bios and tried to surpass 2700MHz NB and could not, so it seems that is where the failure point was. That makes me wonder where in fact the HT bus would give in! I thought these boards weren't supposed to be able to clock HTT much above 250MHz or so?
Next time it goes under DI yes I will do some serious 3d benching, wifes getting me another card to play with for my B-day (feb 5th).Quote:
Originally Posted by Lightman
I won't be so concerned about toasting something If i have a spare.
Unfortunately I'm not in the same financial position I was in 3 years ago when benching so losing hardware hurts a tad more now then it did back then.
Seems like its a strong chip, a better pot might yield some interesting results and LN2 :eek: may be very very interesting.
Not to bad for a gigabyte board "that sux" :ROTF: Boy am I glad I don't listen to "the majority"
I was benching for 4 straight hours with 10LBs DI, good thing I didn't buy 20lbs.
I thought I might add that I think the Myth that you need a HIGH Amp / HIGH Watt PSU is debunked, you just need a quality PSU with Clean voltage. I'm sure if I used a video card I would require more but for aquiring max clocks and cpu intensive benches it makes little to no difference.
damn good run
http://valid.canardpc.com/cache/screenshot/491891.png
thats the best i have gotten on water i'll try for that 5ghz soon hopefully the chip can take it :D
Exactly, especially the point I highlighted though. You'd never max out a high efficiency 500w PSU without a video card or multiple video cards. :)Quote:
Originally Posted by chew*
Stepping? Volts? Temps under load (prime)?Quote:
Originally Posted by cdawall
its all in hereQuote:
Originally Posted by EnJoY
http://www.xtremesystems.org/forums/...d.php?t=215454
If you need any hint's/tip's for 5+ validation DI shoot me a PM.Quote:
Originally Posted by cdawall
BTW all my results were on a bios that Does not support PH II ;)
Updated results in post 2, PI, cinebench, everest 4.0 on air......... TRUE 120 on the way courtesy of a member here, I can finally close my windows and stop wearing a jacket while benching..
How strange is that :confused: a bios supporting Phenom II but lowering OC. So i should flash back to a older bios. By doing that ACC will work again. Uhmmm...... Its worth a shot.Quote:
Originally Posted by chew*
Updating Post 2 with low volt / stability testing 3720 @1.475 already posted.
Next will be 3840 and then higher if capable.....
good clocks at those vcore's mine only does ~3.6ghz 1.45v and 3.9ghz @1.55v
Oh you like that eh? Here's one that won't make the front page for obvious reasons....it was 1.50vQuote:
Originally Posted by cdawall
http://members.cox.net/nfridenti/3.8%201.5v.jpg
3840 stable results up.
3960 is a no go for now with this terrible heatsink for 24/7 stability, it can pass all tests except prime.
Prime heats it up to much even with cold air.
I believe a TRUE 120 is in the mail and on its way
I think that will lower your temps quite a bit :)
Would be nice to see you hit the 4.0 mark :)
yah prime stable 4.0 on air would be sweet, that would top my q6600 extreme air results.Quote:
Originally Posted by Astennu
Atm they are preety matched for max air 24/7 stability. More importatantly I'd like to bench them because once you hit the 4.0 mark I think this chip scales better in 3d....
I might have the golden one (I hope :D).Quote:
Originally Posted by chew*
I didn't try Prime yet, but that setting was for about 1h benching with no problems.
http://img150.imageshack.us/img150/9...zcinepivw9.png
All thanks to 32bit Windows ;)
BTW chip was Pi stable @1.475V and Cinebench 2003 full set, but R10 killed it half way through multi-core render .
If tomorrow I've got more time I will try to Prime it.
Ps. Validation @4.2GHz 1.58V
http://valid.canardpc.com/cache/banner/494259.png
Looks like it scales nearly identically to my chip in terms of total cpu overclock per volt. :up:Quote:
Originally Posted by Lightman
http://img.techpowerup.org/090127/Capture026.jpg
anyone else seen sciencemark it takes ~5min to run and gives a pretty good run between AMD and intel
New bios I'm working on now, F3o from gigabytes site, solved my questionable PI scores, 32m 1m 16s faster at same 3840 mhz speed.....post # 2 last 2 of pics;)