|
|
21-12-2008, 03:33 PM
#43 (permalink)
|
Originally Posted by CryptiK
Did you manage to do any higher than this yet? I was hoping this board has more in it, as I max out at around 542 FSB / 650 MHz on my P5Q-Deluxe.
Also regarding the max FSB of 571, managed to push this any higher?
After looking carefully at the BIOS options, I was wondering what the effect of altering the MCH Ref values is? I know with for example the Gigabyte UD3P the MCH Ref setting is critical to get high FSB. I assumed you tried playing with them, but if you didn't perhaps it may allow it to go a little higher. It'd be nice to see 580 - 600+ out of this board, as boards like the Rampage Formula can manage to do it.
Lastly, how hot does the digital PWM get? I've heard conflicting observations about digital PWM, some asy it runs cool, others say it runs very hot. Is the cooling for it sufficient?
|
max 32M pi stable FSB i last tested was 563FSB - haven't tried pushing higher as yet as i seem to hit a 5:6 divider mem clock limit around 658mhz for super pi 32m stable settings so far. My aim is on fast 32m pi capable FSB so 1:1 doesn't cut it so far heh. Have yet to try 1:1 max fsb with 4-4-4-x timings.
MCH VREF 1/2/3 you mean - i did play with them only briefly - alot of values just end up with system no boot needing clear cmos, only value that had a minor effect in stabilising high mem clocks was MCH VREF 3 at value of 25.
Digital PWM is rated for something like 110+ C, but it runs relative cool but i have a fan over than area anyway
Originally Posted by GenAn
There are quite a few permutations that can be derived from those four settings with all their options. I thought I found a DFI PDF that had suggested settings for certain DRAM chips or FSB settings but darned if I can find it. Getting this board to recover from a bad OC is a lot harder than the LT P35 where you just held down Reset and Power for 4 seconds.
|
Failed oc' recovery for me is shut down rig wait a few minutes and power up rig hitting the home key or delete key and loading last bootable settings from CMOS Reloaded
Originally Posted by GenAn
I guess the other thing I wanted to know eva, is if you had any kind of system for testing the DRAM clock/driving options and skew models. There is scarce little documentation on what they actually do or even which levels are high and which are low.
Oh, and if you could share some of those BIOS settings for your high 2:3 memory clocks
|
memtest86+ is my starting point to see what each combination has to bring for max memtest86+ stable mem clocks, then i move on to windows tests. I always follow same approach as outlined at
http://i4memory.com/f80/announcement...y-eva2000.html and the linked memtest86+ approach
http://i4memory.com/f79/memtest86-v2-10-a-6/. There's no need to know which are high and low levels for what skew models are for each modules as i just go by what works for the particular memory I'm using on particular divider/fsb/mem config combination
Originally Posted by CryptiK
Just another note, 3oh6 says Rampage Extreme uses the ICS 9LPRS919BKL clock gen, which is the same clock gen used on the P5Q-Deluxe, whereas eternal fantasy says its ICS 9LPRS918JKLF. Wonder who is right or if maybe both were used at different times/revisions.
Does increasing/altering the clock gen voltage and CPU clock amplitude affect/increase the max FSB achievable with the T2RSB Plus?
|
Probably a revision occured ?
Yeah clock gen voltage and cpu clock amplitude can stabilise FSB on the higher end borderline stable mark but it isn't a big boost.
