Yeh, me too, I think this bios sets tighter tRFC than any other bios. I'm thinking I'm going back to 1401..Quote:
Originally Posted by Mykou
Printable View
Yeh, me too, I think this bios sets tighter tRFC than any other bios. I'm thinking I'm going back to 1401..Quote:
Originally Posted by Mykou
Why don`t you ask eva2000 ? He has direct contact with asus :) .Quote:
Originally Posted by lawrywild
would it be out of line to ask him why vdroop is intentionally present on all ASUS boards?
AFAIK, they're working on such a bios hehQuote:
Originally Posted by Pjoeloe
Eva knows about the tRFC thingQuote:
Originally Posted by Pjoeloe
edit: oops too late :p
There are definitely some different memory settings going on with this bios. My week 27 6600 does 3.3ghz default volts, & 3.6ghz with 1.475v set in bios. I have to set vcore over 1.525v get it to boot @ 3.6ghz. I just tried 445 FSB to see if it was the ram or something & it failed 8 MB Pi. Loaded memset changed timings, it passed 8 MB Pi.
Yeh same here, it seems to be setting different tRFC at certain dividers/timings than other bios revisions :stick:Quote:
Originally Posted by The Nemesis
The power down feature that allows the overclocks to work is no longer functioning properly. I think thats why I'm having trouble overclocking with this bios. That said this bios is still very promising for when its finalized. Its nice to see that asus is trying to get the P5W DH options equal to the P5B. I was not going to get the P5B because 8X sli is a must for me. I still might try abits new board if asus doesn't come through with a good bios soon.
When you say its "no longer functioning properly" what do you mean? Do you mean it doesn't power down now (cold boot) when you do a reset?Quote:
Originally Posted by The Nemesis
If it doesn't that's good news surely?
Does cooperate with SpeedStep? If I set it to a lower multi, will it artificially cap the multi at the one I've set it at, instead of reverting to the default multiplier?
Quote:
Originally Posted by Durzel
Yeah, good news for the Vapo users. :banana:
The Nemesis please confirm, thanks
Nice bios... but i think, the other one was a bit more stable at the same settings... :slapass:
with my p5w, both 1401 and 1402 will make pcie to crap out at 1.9v vmch whereas with 1305 and earlier ones it was good 'til ~1.98v.
as i have currently running at 1.91v i'll skip this bios :) as i have 6400, don't have any features to miss.
another thing - with 1305 i have trc (bank cycle time) 16 in cpu-z, whereas with 1402 i had 8. all memset timings stayed exactly as they are.
edit: i have to recheck the last bit because trc = tras + trp and should've been the same with both bioses...
hum, this bios is very buggy with my rig, i get issues running futuremark and made all my benchmarks messyQuote:
Originally Posted by lawrywild
lost about 4000 pts in 3dmark05 and freezes the whole system , very unstable as never before !! :mad:
tRD / read delay - shut down after saving bios boot problem is fixed for the P5WDG2-WS PRO see http://www.xtremesystems.org/forums/...3&postcount=42 :) .Quote:
Originally Posted by eva2000
Now the Row refresh cycle time (tRFC) :p:
For me its not doing the power down after setting bios overclocks. Its not doing it after a restart from windows either. Its not consistent, there were a few times it did power down. I finally stopped overclocking with it because of the weird or corupt bios screens I was getting.Quote:
Originally Posted by Durzel
Sounds like a pretty buggy BIOS, but then I suppose it is beta.Quote:
Originally Posted by The Nemesis
At least enabling multiplier adjustment in the BIOS is a step in the right direction. Shame it's not available only to ES chips though, wondering why I paid out the extra now ;) (joke!)
Was the OC Profile working prior to this bios. I haven't tried it since the 1101. Works fine on this bios though. May go back to trying some overclocks now.
TONY have a good explanation :
Quote:
All 975 boards have issues above 350fsb 1:1 due to the nature of the strap settings and memory ratio's. What I am saying is for Abit to stand a chance they HAVE to implement the 1333 option in bios for you to choose at 350ish and this will allow the board to sale thru 350fsb to well over 400+fsb. The issue is this though, ram performance will be slower. Now i asked abit to add the NB latency settings into bios also with the hope that the 1333 slowness could be countered by altering the latency, pretty much like you saw on the IC7 boards.
Micron based dimms are more tollerant and give the memory controller an easier time...they will allow 1;1 and above 350fsb easier using the 1067 strap, but they will still top out.
ALL the 975 boards at this time excpet badaxe are using the 1067 strap and not moving to the 1333..this is the reason they are ALL clocking less than 965. 965 is also a better clocking chipset and will clock higher in the strap range than 975 but it does still top out, this is why we see asus using 1333 at 400 and up.
So in short the dimms may limit you a little yes, the nature of the chipset is also limiting you and unavailable bios options
http://www.xtremesystems.org/forums/...=115839&page=4
I think I'll stick with my 400x9.
That would explain a lot! :)Quote:
Originally Posted by mascaras
Is Tony saying that the Asus 965 DOES switch to 1333 at 400? Or doesn't? Didn't he say only the Badaxe did it properly? Not quite sure I understand specifically what he's saying.
:rolleyes: I saw in another thread the new P5W64-WS PRO also beeing able to lower the multi. So does this BIOS also work for the WS series?
I'm using the P5WDG2-WS PRO and I'm still not sure to flash or not.. :confused:
Someone already tried ???
Dont even try at best it just wont let you flash it, worst you will have a dead boad.Quote:
Originally Posted by camouflage
Just wait until you get a new bios rev for your board.
New bios 1403 Beta
http://taskforce.no-ip.info/pics/ASU.../p5wdh1403.zip
And
MemSet for 3.0 version:
Memset 3.0
If I resume changes between 2.1:
-add support for Intel 965 ,AMD A64 ddr&ddr2;
-not necessary to install/uninstall driver;
-Save function was simplified.
-finally all the code was changed...