Well, I don't think this board is that bad. I think that almost every brand new board I have had, have had some BIOS-issue.
Printable View
Well, I don't think this board is that bad. I think that almost every brand new board I have had, have had some BIOS-issue.
Never seen a board that would flat out not work at stock settings for an entire line of RAM chips [micron]. Someone really dropped the ball there, I think that only a handful of board makers can have this happen and not lose a ton of customers.Quote:
Originally Posted by blossa
I don't think this board is such a bad one either, take a look in the other board threads in this forum - ppl are experiencing same type or other problems with every new board. Take a look at the new X48 DFI board for instance. I've bought several DFI boards in the past as soon as they were released, some of them had problems for months after it finally got fixed. I find it irritating that the industry works like this, but I guess we have our self to blame as we all want the cutting edge technology right away...
Kind regards,
Karl
ASUS has three sets of the OCZ Platinum 3P1600EB4GK kits, all three exhibit different characteristics on this board, two of the kits have different specs but the same part number. I ordered another set from NewEgg to try, same held true here for my first Corsair kit (to be fair) as one had the same problems as the OCZ, but the second one is humming along at 1600 easily. I am also swapping out my GSkill and Patriot kits this week for newer kits. This regression testing is getting expensive... :DQuote:
Originally Posted by GAR
Kinda looks like a IC or SPD change to me at first glance but it appears to go much deeper now. That said, getting a BIOS fix is proving difficult due to all the changes, one memory set works great, the next one does not, get the one that does not work up to speed, and another kit all of a sudden has problems. By the way, two other P45 DDR3 boards I have are experiencing the same problems, so it might be a core BIOS code problem.
You are right to blame us for part of the problem; why would they pay for R&D when they get free guinea pigs?Quote:
Originally Posted by Karl
I had two sets which acted the same.....they all use Micron, they should be acting the same, maybe its the board acting different??? If it was an spd issue, one was working, and the other was not, i think tony would know about it.Quote:
Originally Posted by bingo13
I have six Micron kits here, two work perfectly, the other four do not. The "bad" four ( :D ) also do not work on the MSI Platinum board but the two good ones do. It is the boards, BIOS, or IC/SPD combination? :shrug: That is what we are working on right now. As a side note, the Samsung kits work perfectly regardless of brand.Quote:
Originally Posted by GAR
What about bios 0804 is it doing thomthing or not??? , whre can we get it ?
You can't get the BIOS 0804 yet mevorach.
It's not perfect, and now I have some symptoms that could be sourcing from the BIOS ( every BIOS has a/some flaws usually ).
Give it some time, I bet we'll have some Micron DDR3-1600+ 4GB Kits action soon.
Last night, I testing about 2 hours with 0801 bios. I tried how performance p45..only 5:6 ratio worked well for me. (2GB OCZ 1600EB kit).
I got now 0804 bios, and I go test again..
http://img184.imageshack.us/img184/4474/11128bi0.jpg
it seems 1:2 ratio works. i could reach DDR3 1800 but at 2000 BSOD.
http://img371.imageshack.us/img371/8...0c7sul3.th.jpg
I'm having trouble with 2x1gb kit of Cellshocks. Those are Micron ICs too I believe? At 1600 mhz, spec timings and even with Tony's suggested relaxed settings memtest spews errors like crazy. I know the sticks are good, they worked fine up to 1800mhz and a little beyond in P5E3 Premium. Still on bios 704, I'll flash that to the newer version as soon as I get home tonight and see if that helps. :(
A little update to where I am:
been testing multiple kits, all 1 module at a time. Some work right off the bat close to 1600 with little to no skew adjust needed. Some work with a little skew needed and 5 mins fine tuning gets you rocking along 1600MHZ single channel.
Now here is where it gets complicated:
Thru testing A2 and B2 (orangle slots as Asus recomend) I have found the following.
1 I can dial a module in A2, for instance 500ps NB clock delay, A1 skew 250 advanced and A2 skew 250Advanced
2 remove that dimm, place it in B2 and try to dial in the skew for B2 using only B1 and B2 Dram clock skew adjust, I have to leave the NB skew good for A2 ;)
3 It will not dial in, nothing I do will get me stable with the same dimm I used for A2 stable in B2 leaving NB skew at 500ps delay.
4 So, i move to change NB skew and find B2 200ps delay Dram clock and Normal NB clock gets me the dimm stable at 1600MHZ.
5 Now move the module back to A2 and try to dial it in again....no joy :(
6 Try leaving B2 module alone and use one of 6 other modules I have here leaving the B2 NB skew and Dram clock skew intact and tuned....Nothing works, A2 needs a different NB skew to B2
So Gary says he has kits that work..2 from 6, I have 6 kits and I can get any of the modules tuned in but ONLY on their own.
I asked for a bios that has DRam CLock skew, DRam DQ skew, and DRam DQS skew and labelled as such. I feel I will stand a much better chance dialing in modules if we have these 3 skews all sperate. As it stands now Asus need to work a few wonders in a bios ( which I am sure they can do) but i really would like to do my own tweaking with this to be sure I understand what is going on.
it is SPI 32M stable with 1.96vdimm. my RAM has GTR chips.Quote:
Originally Posted by jugeen
http://img395.imageshack.us/img395/5...i32mzn8.th.jpg
Anyone has graphics interface link width 2x issue with PCIe graphics card?
jugeen i think 1:2 worked nice on 0801, i reported that few pages before :)
What kind of PCIe problems you faced ?
ASUS P5Q3 Deluxe BIOS build 0804
This one improves compatibility with Micron based 2x2 GB modules (especially OCZ Platinum or Platinum EB PC12800)
Improves does not mean - it fix it absolutely - so please don't complain at it atm!
sry for re posting. so my problem is that i cannot fix link width 2x anyway... :confused:Quote:
Originally Posted by pmp
I think it needs to see the ocz part number, so if you have anything different than EB you may see no gains.Quote:
Originally Posted by pmp
Thats a "may" Im not 100% sure but it was created with 1600EB on the bench
I tried 0804, everything same.. my 2*1GB OCZ Platinium 1600EB kit not stable at 1600 or more...Quote:
Originally Posted by pmp
But this kit could 32M stable results in Asus P5E64WS EVO :
1600Mhz 6-6-6-20 2T 1,98v
1800Mhz 7-7-7-20 2T 1,98v
2000Mhz 8-8-8-24 2T 1,98v
Tony: I think SPD is no good for P5Q3, because last time my rams no boot in nf790i too.. What do you advice? What SPD I flash for my kit?
There is no specific spd at this time, DDR3 is a different beast to ddr2, with DDR2 you tell the board what to boot at and with...DDR3 gives the board a set of calculations it has to do...then it boots.Quote:
Originally Posted by over@locker886
Im workin on it for sure, I have a 2GB test spd that seems to make a difference, BUT im not releasing just yet, plus P45 has a new smbus that w1zzrad needs to code into spdtool.
all my worries has flown away. my problem was graphics card related. i swapped it to a 9800GTX and after a GTX280. with both cards link width was 16x. so there is no PCIe bug like we had before with other Asus mobos. unfortunately my 8800gts seems to be faulty cause link width 2x error appeared on P5E3 Premium and P5E64 Evolution mobos. it's RMA time. :(Quote:
Originally Posted by jugeen
Thanks, I will try it later today.
Short Update -
Intel is busily updating the core BIOS code on a daily basis now in order to solve the DDR3 problems. More later...
Get my board tomorrow. My Kingston uses Samsung chips. I will update.
Looking forward to this. Thanks:DQuote:
Originally Posted by bingo13