Looking over at the Rampage Formula thread, I wonder if we can't get something similar going here to request for DFI to make a new BIOS for the LT/UT P35 with added options like drive strength, DLL Skew Models and CPU clock skewing.
Looking over at the Rampage Formula thread, I wonder if we can't get something similar going here to request for DFI to make a new BIOS for the LT/UT P35 with added options like drive strength, DLL Skew Models and CPU clock skewing.
zsamz,i understand that i have to play with them and i understand the chart are a base but my question was what are the formula for calculating the values. like in this thread?
http://www.xtremesystems.org/forums/...d.php?t=202292.
zsamz how did you come to this %?
http://www.xtremesystems.org/forums/...postcount=4992
does anyone understand what i am asking?what is throwing me off is when you turn on the gtl's it start at 110/110/110.here is an example,say i set 1.30vtt and i set 78/78/68 and then someone ask me what percenage of vtt i set my cpu 0/2 and 0/3 to.and what % i set my NB gtl's to.how would i calculate it on this board so i could answer them? i can do this with the UD3P and the asus MIIF and the P5Q-deluxe.but no one seems to have an answer for this question pertaining to this board.they just say"you have to mess with them to find the right setting" or "here is a chart" or "get your DMM out and test it to find the right value".i understand all this but again that is not what im asking?
they put drive strength in an x48 bios then they removed it for fear of users will kill ramQuote:
Originally Posted by Shiranui Gen-An
skews would be nice but i doubt they gonna do it since they rather sell you a ut p45
i just use it as a baseQuote:
Originally Posted by radaja
the gtl setting that gets me stable @ the lowest cpu volts is best for me
gtl settings dont need to be exact with a dual core
but with a quad it needs to be more precise
again im not asking how to adjust gtl's,i know very well how that is done.i am asking how are gtl values calculated on this particular board,using the actual values that the board gives(110/110/110) and calculating them into percentages(%63/%67/%67)of vtt.
Ok radaja, I think what you are looking for are tables like these --
EdgeOfStability - CPU & NB GTL Default Settings
Anandtech review of DFI UT P35-T2R - page 12: Quad-core CPU VTT/GTL Tables (do continue to read pp.13 & pp.14)
HOWEVER, the figures of those tables were sampled from just a couple of boards back then. Keep in mind that your voltages may vary! That's why it's always recommended that you measure your own board to get a much better idea. Although, those tables should serve a good starting point for you, who come from a different calculation methodology. And do study what the BIOS settings translate into as they don't mean direct % or absolute voltage values. There will be a lot of try and error so you need to be a little persistent to master your own board. Good luck.
I think we need updated tables for 45nm CPUS, those are all for 65nm.
well i wasnt able to lower the volts very much (one notch vcore and vdimm) but its good enough for me and its stable.
http://i44.tinypic.com/swa461.jpg
http://i44.tinypic.com/rswk8x.jpg
cmmon radaja you can do better
off topic for this question. i started having strange post codes before the usual "C1", like "1C" "1U" and more often ".88.". i spoke with dfi tech support and was informed that this could be a corrupt bios or a bad power supply and ".88." is an LED segment check. i noticed these post codes after putting in a new corsair 850tx but the codes could have started happening before this new PSU. i tried a PSU from a different system, where the same LP would go to "C1" and no other post codes would show before the "C1". the PSU from the other system would allow my system to boot a normal "C1". so i put the corsair 850tx in the other system and had the same unusual post codes before the normal "C1". today i got a replacement antec PSU and am getting the same strange post codes before the "C1". ??? does it appear that i need to RMA the new power supplies, corsair and antec, can i fix this with a bios flash, or do i need to RMA my motherboard? thanks in advance. :confused:
You must have gotten a golden board Radaja, I don't know how you're running PL7@540FSB and memory at 650MHz with your vMCH at 1.50V.
Is performance level at same fbs and ram divider equal for p35 and p45? Because it seems that many p45 have dafault (auto) pl at > 9...
The auto performance level for this board seems very aggressive to me (so you have to raise the vnb to gain fsb mhz)
From what I've seen, P35 can run PL one level lower than P45 can at the same FSB and memory strap/divider.Quote:
Originally Posted by litmanenyari
Yes it does run the PL one level faster than the P45. The downside of that off course is that it will not clock the FSB as high as the P45, or clock the FSB as high with all 4 ramslots populated.Quote:
Originally Posted by Shiranui Gen-An
By the way, PL at 7 is the defult PL when running on the 5:6 divider with the P35.
any one else using bios 9/19? i had to flash to it today again now that ive installed my 9650 cause after two fresh installs xp and vista in device manager under processor it just showed unknown device.but with bios 9/19 it works fine.only problem is that everest wont show nb and pwm temps or voltages such as vtt vnb.any one else experience this?
yup thats why i went backQuote:
Originally Posted by radaja
i like voltage readings
you can use dfi proggy though
yep,i guess its better to be able to read temps and volt than to have cpuz say Q9650.right now im on 5/02 again and it all good exceptdevice manager just say intel cpu and cpuz and everest show intel processor 3.00Mhz instead of Q9650.but at least its better after flashing then reflashing bios cause now it intel processor and not unknown device.
http://i41.tinypic.com/sowh7p.jpg
http://i41.tinypic.com/2vhz0ck.jpg
its quite common for dfi board and everest to have conflictsQuote:
Originally Posted by radaja
i had that problem every time the x48 got a new bios lol till the lavalys team updated everest
if you really need it you can ask them 2 everest forums to fix the issue
has anyone succesfully gotten a 45nm quad core to boot at 500fsb with this board?if so could you post what was needed(vNB,vtt)
I think the lack of a CPU Clock Skew option is hindering the board with the Yorkfields and 500FSB.
never tried with my xeon, but i managed to do this with my 8400 and last bios, before packaging the board and ship it to the new owner.Quote:
Originally Posted by radaja
http://forum.hwproject.net/imagehost...a77b3aa70b.jpg
see also my sig :p:
and yes, i'm 7ornado from i4memory ;)
yea right when i saw the screenshot i could tell it was you.heres where im at now.still priming ,we'll see how it goes.P.S thanks for the bios settings at i4memory.
http://i41.tinypic.com/16h5bw0.jpg
:shrug:
EDIT: for those of you trying to find out the meaning of DFI P35 post codes including the infamous 88 try Clunk's.:yepp:
thanks for the help. :DCode:http://www.clunk.org.uk/forums/hardware/3203-dfi-bios-guide-intel-chipset-based-motherboards.html#post12434
Quote:
Originally Posted by the9093
Hey Radaja
I just flipped out my q6600 @ 3600 for 9650 @ 3600 , now im trying to push it
to 4.0 at a vcore of 1.300 vtt of 1.3400 nbv to 1.52 and when i set the gtl ref to disabled " board will boot but fails quick on prime , if i set to 66 66 66 66 board wont even boot , cant you help me a bit wit voltage settings
ok I see a lil voltage problem here , when i set it vcore to 1.360 it comes in on cpu -z as 1.322 to 1.344 and fluxes back and forth , im prime at this then start lowering down my vtt , wish i could figure out the gtl s then im sure i could bring the voltage down , should i be using special add voltage , i never needed with my q6600 , man for a low vid 1.2500v this sucker needs alot of juice and runs a bit warm in prime 40 c but it still runs like 10 -14c cooler than my q6600 , damnnn core 2 just failed after 10 min
core speed at 4052 9 x 452 at 1/1
bios 9/19
CPU vcore: 1.360 1.370 in smart guardian
CPU special add: auto
NB voltage: 1.45V 1.42 in smart guardian
MEM: 2.1V 2.06 in smart guardian
CPU PLL: 1.55v
Clockgen: 3.45V
CPU VTT: 1.40V
Vcore droop control: disabled
GTL values: are something like this Core0/Core1/NB : 110/110/110
did you reflash biosQuote:
Originally Posted by the9093
c1 is normal if you got memory problems
88 happens if you got a bad overclock