I am one of 5 Asus Motherboard Moderators that has a connection straight to the Taiwans BIOS Department. And what Fixer says is correct.
http://vip.asus.com/forum/profile_vi...ge=1&count=244
Printable View
I am one of 5 Asus Motherboard Moderators that has a connection straight to the Taiwans BIOS Department. And what Fixer says is correct.
http://vip.asus.com/forum/profile_vi...ge=1&count=244
Kensek,
Please update us if there are new bios fix the c1 quad core high fsb issue !
I am now limited to 450x8 1.24v at P5K-E.... MyQ9450 should be able to go further...:)
It's all about voltages. A CPU FSB of 400 usually requires 1.63v on the NB. 450 FSB 1.67-1.70v. 475FSB 1.70v+
Also the CPU needs 1.325 for 450FSB. and with VDroop 1.355v.
If your your BIOS doesn't have the full barrage of Voltage adjustments you are screwed, more than likely.
You need to tweak NB to 1.75v, CPU Pll to 1.75v, FSB vtt to 1.50v, CPU to 1.45v, sb to 1.175v, etc.
Isnt CPU PLL and VTT voltages a little excessive?
Look at my Voltages when my X3350 is running at 480x8 on my Rampage. They are shown previously in this thread.
Anyone looking to run a CPU FSB >450 with a NB lower than 1.60v is living in a Fantasy world.
Hmmm...?
3.6Ghz with 1.25v NB here on a P5K-E (maybe just a nice chip/mobo fluke?):
http://www.xtremesystems.org/forums/...&postcount=333
Just tested the Q6600, 475*6 is PRIMEABLE & stable with Same P5K-E wifi pcb1.03
vCPU 1.325v
vFSB 1.3v,
vCPU PLL 1.6v,
vNB 1.4v
Load line calibration enabled
Same P5K-E wifi pcb1.03 (P5K Premium pcb) For the Q9450 it is primeable 6hrs large fft at 450*8 at bios 1.2625v CPU, vNB 1.25v, vFSB 1.3v vCPU PLL 1.6v. But at 475x8, I set all highest voltage with active cooling fan blowing.... Still cannot load in to windows. System hangs !
So, 2 answers: my CPU fsb wall or the bios issue !
Btw, the vNB 1.25v is me !
this is madness!! but seriously you'll probably burn up your mobo with 1.6v+, +1 with mach82.
it takes 1.4v for 450fsb, with 4dimms 1:1, Plvl 2 (=8) on my P5K-E. The only reason i cannot go lower is because 1.25v on the NB bugs out and is actually 1.55v.
If you can run those low Voltages on a Rampage or a P5E3 Premium, then I'm doing something wrong.
My P5E3 premium couldn't do anything higher than 450x8.
The maximus did only 440x8.
The Rampage, I'll blow that mother up. But it runs super cool. NB SB at 30C at those voltages on Air.
Anybody know what the Phase pull down thingy does in the Rampage bios?
That's seriously too high. Maybe you should try loosening RAM timing, or lowering RAM frequency to 1:1. On my ASUS P5E-VM HDMI, 450FSB requires only 1.25v or 1.35v on the NB with the IGP. 475 requires 1.41v, or 1.51v with the IGP, and 500 (stable) requires 1.55v, or 1.65v with the IGP.
i failed to go above 450fsb for my x3350 on p5e3 premium with bios 0401. then i flashed to 0204 bios and i manage to hit till 473.
here is the setting i use:
vcore 1.36
cpu gtl 0.66 & 0.67
vtt 1.36
vdimm 1.8
vnb 1.41
nb gtl 0.67
sb auto
it might be not prime95 stable, but multiple PI 32m is no problem..
but then the 0204 bios is not good enuf, i failed to enter my vista few times. i thought my oc screwed up my os but then i realize it was bios issue.
i flashed back to 0401 and guess who's back again.... the 450 fsb wall :(
so it's not about the voltage settings, i'm pretty sure most of us here is experience enuff to test it step by step and to know whether it is insufficient voltage or other matters that cause the oc to fail...
Still waiting on my X3350 to ship..... :/
My P5K seems weak though, cant seem to push the PL at all and with an E4300 333fsb is absolute tops..
http://www.xbitlabs.com/articles/cpu...0_6.html#sect0
Look at the link above. They tested a Q9300 at 467x7.5. NB = 1.65v and CPU pll was 2.00V.
I sold all of my X3350's now, so I can't play around with them.
Might get a Q9650 when they come out in 2 months for $540.
Or I may try another QX9650, but my first one couldn't even do 4.0GHz - so I sold it.
All I want to do is to find a Quad that can do 4.0GHz on air.
My Old Q6600 could do 433x9 on Air for 3.9GHz. Therefore I thought a Q6700 with a 10 multi would get me there. So I spent $275 for a new retail Q6700. What a disappointment - it can only run up to 375 x 10.
Or so we all hope at this point. Paying $300 for a motherboard, only to be beaten by a $120 board can be quite frustrating sometimes. Heck, the chipset on the cheaper board doesn't even look like it was meant for extreme overclocks. (and yet it still did some extreme clocks)
yeah, i hope ASUS can open their eyes to accept the truth that it is the bios that need to be tweaked.. not the setting :P
Kensek - Since you've got contacts in the BIOS department can you have them look into whats different between the chipset strap/timings of the P5K and P5E?
OK I E-mailed my contact your question. All I can tell you from my past inquireries is that each board has its own BIOS department.
And once a board gets so old the BIOS support is no longer active on that board but has been reassigned to the newest model.
And when I'm talking Old, I am not talking about years, but months.
Thats why in the 8 months I've had a P5K Deluxe, a Maximus Formula, a Rampage Formula and a P5E3 Premium.
P5K support is All but gone now and so is most of the X38 support.
rite now so many boards has been produce but does the engineer also become more efficient in their work?
when u say OLD is just for months, then that's why sometime they just leave things unfixed.. bcoz they already move to the next board..
board with has so potential, but spoil by human who works inefficiently..
sorry if my words are hash, but when u buy sthing that cost u almost of ur salary (the price for p5e3 premium in my place is RM1500, while average salary for fresh graduate is only RM2000) then u expect it to perform good, with fast and good supports from the manufacturer..