some more testings:
http://images0.hiboox.com/images/400...0aea1cb870.jpg
http://images0.hiboox.com/images/400...197cd62176.jpg
Printable View
nice results but ehm... is it better with the new bios or not?
eh hem... sorry, I forgot to tell, but yes it looks easier to reach high fsb. I have to test if it is not slower with this 0605 than with 0505.Quote:
Originally Posted by JohnnyV
Does anyone have any information on the new 0605 bios? Is this a beta bios or an official release? I ask this question because the 0605 bios is not on Asus's website. Any information on whats new in the 0605 bios? Thank you.
I think it alows more fsb and this in benchs, not sure anywayQuote:
Originally Posted by deachus
More FSB for dual cores or for quads? There has been a FSB wall for my Q9450 at 495FSB. I cannot get 500FSB bench stable so I am hoping this bios has fixed that problem.Quote:
Originally Posted by ToyTen
Will test when I get home tonight.
dual in my case
Quote:
Originally Posted by deachus
It's not official. If you follow the link to the post with all the new biosses you will see who posted them and that it's a beta release.
its a beta i think , not tested my self but will tomorrow orso .
650fsb screen last time, 615fsb cinebench, 620fsb 3d and 645fsb pi1m so i hope it will bump that up a bit :D .
toyten: any advance on ram clocking ?
Quote:
Originally Posted by blind_ripper
Those fsb's on AIR for the NB? I quess but it must be on the limit :D GG !
looks more flexible (I'm still using a dualcore), better.
But can't tell if it is more nervous at the moment.
all air , think the 650fsb screen whas with cascade on cpu .Quote:
Originally Posted by ToyTen
no diffrence to air thoughe so it whase the board , still i did not tray to tweak the board :D .
mem clocking whas a mess , only could do 950mhz cas7 on D9JNL so yeah i hope the bios helps on this .
no i need to get a other 3d card killed the 4870 when putting on stock cooler :D .
so my 642 isn't that bad, cpu under water :p: ?
'Should put my autocascade on it,time won't last to come :)
I have GTR mem's so can't compare, I also have JNM(G-Skill 12800 7-6-6-18) and those sucks...
I couldn't bench with GTR at 1Ghz CAS7 so easily with 0505 BIOS so thats a good thing yet :)
I found my QX9770 didn't like 500 either, but 501 worked just fine. As I was going up, I had the same issue at 450. It would not stabalize there at all, but when I went 451 everythind was stable.Quote:
Originally Posted by deachus
I am finding this board very finicky with the quads & realize small steps got me furhter than large ones.
Hope that helps you out.
This is my first attempt OC'ing my new CPU:
http://img143.imageshack.us/img143/9...1840da2.th.jpghttp://img143.imageshack.us/images/thpix.gif
I know its still high on voltage but I will try to reach 4.5Ghz later. After I'm done doing some benchmarks on 4.5Ghz I will determine what my 24/7 clock will be. Then I will start finetuning vcore.
Well after alot of testing I am loving the new 0605 bios. I am now able to clock past my 495FSB wall. I was able to take the number one spot for 32M on HWbot for the Q9450 and get in the top 10 in almost every benchmark. I am going to run some 3D tests and see if I can't do better on SuperPI32M. Here are some pics.
What vNB are you guys using when having such high FSB's? I found my CPU was stable for 9 hours of orthos but when I also added higher mem speeds and set vNB to 1.61v it only was stable for 1 hour. Is it reasonable to put it any higher then that?
I used 1,85vNB to finish my bandwith at 642 fsb.
http://i3.photobucket.com/albums/y90...DSC00011-1.jpg
socket A cpu cooler on the chipset right now :D .
Cool, well done and looks good. How does it perform?
full stock base sink only gets 27.2°C , with 1.7vnb it get 33°C so no wurry's :) .
ready for bios 6050 :D , now il need to change cpu again :p:
I used vNB 1.81v for 503-505 FSB. I still have the stock NB HS on their with an 80MM fan blowing directly onto the NB hs and it's just barely warm to the touch, not hot at all. This setup is feeling extremely fast.
I did some more testing late last night with the 5:8 mem divider running the mems at 1617Mhz @ 6-5-5-15-1T tRC 62 505FSB and this was the fastest my setup has ever been. With these settings I was able to score 11m31s in SuperPI 32M, which beats my Q9450 32M world record by more than 12 seconds but once I opened up CPUZ a couple times and took the screenshot and pasted into paint the computer locked up before I could click save. I was NOT happy. It's Friday so I will be tweaking the hell out of this bios this weekend.
@ Blind_Ripper
It sounds like it's worth it to mod the NB HS that way. Was it easy to mount, did the mounting bracket fit in de old NB mobo holes?
yeah i hope so :), extra testing : 1.7vnb = 35.8°C idle and 36.3°C load :) .Quote:
Originally Posted by Pisklink
reading with a probe at base off the bloks .
well try WC and see how that give's any advance .
mounting easy offcourse , drilled holes and bolts treu it .
Just started benching again and beat both my 1M and 32M records. The 32M beats my world record for Q9450 on HWbot by about 9sec.
Back to benching for now:D
Next quote is coming from the E8600 OC thread but we want to keep that thread clean of P5E3 talk so I copied a quote here.
Well I used a 5:8 divider first to make sure I used some decent RAM speed already while getting my CPU stable. Now I got my CPU stable settings for 4.4Ghz I decided to add more RAM speed using the 1:2 divider.Quote:
Originally Posted by Infa
RAM settings with 7-7-7-20-2T @ 1860 with PL6 and 2.0v is all to tight. Therefore I tried PL7, 8 and 9 and didn't pass Prime with it. Because I didn't know yet what a save voltage would be to try I decided to put everything on auto setting (9-9-9-24-1T) and just try if it would pass Prime on blend modus with 1860Mhz. Well at this moment it's running for 3,5 hours. If it gets around 6 - 8 hours I will try 8-8-8-20-1T and see if it passes. If it does I could still try CL 7 and add more voltage to see if that helps. Before I started Prime I did a quick Everest benchmark and already got 11k+ read and ~10k write and 45ns latency. I think that's due to the 1T setting which my RAM won't accept when running at CL7 and 2.0v.