Printable View
1.fix 6/15 usb bug
2.The 1st power on after AC OFF
ac off , not off clear cmos --> "program dram voltage based on user select value(only up to 3.2V) "--> detect dram --> testing the dram --> program the dram voltage up to the final value if you select higher than 3.2V
6/18-1
6/18-2
6/18-3
If you can not oc as high as previous build , report the detail of your situation ... this build will be the release candidate if there is not other big problem reported ...
THANKS OSKAR :)
What's the USB bug?
Just like b4 ????
6/18-1 general purpose (most ram)
6/18-2 UTT/BH
6/18-3 TCCD
Edit: Just flashed from 510-3 to 618-3 and no issues, usb is working fine, the only thing that I noticed was that in 618 the NV Lan Mac defaults to MII in bios instead of Pin strap in previous bios revisions. this may cause your lan to not function until you change the settings in bios.
There was/is an issue with using USB keyboard with the 615 bios.Quote:
Originally Posted by exscape
No USB issues for me so I'll wait until Pershoot adds this to his bootable CD before trying :)
Thanks for the info WWMe :)
TCCD or UTTQuote:
Originally Posted by Major
I like -1
oskar tell me
-1 from 0326
-2 form 0316
-3 form 0310
not dram
Then it's like before, 0316 is the UTT/BH-5. Will add it to my website.
Quote:
Originally Posted by Major
wanna know too...
Try to load optimized default , what I see is the optimized default is still pin strap not MII ...Quote:
Originally Posted by Major
And Try to re-enter the MAC adress if you lose internet connection after you flash bios ... You may clear the MAC address and cause the internet problem ...
Just flashed to this BIOS, changed back to Pin Strap manually (whatever that means, just as long as my net works in happy!). Everything else working fine.
any OC improvement? more stability ?
Yay! I hope this fixes my USB mouse, PS/2 keyboard problem. I've been on the 414's because of this problem in both the 510's and the 615's. Lets give this a good go.
Dont seem to be able to OC further, TwinMOS SP BH-5 still wont do more than DDR500 2-2-2-8 stable at 3.5 volts, and I'm not going to throw more power at it untill I'm done with uni and needing my PC for work. The multi's still have problems, 250x10 still requires more voltage than 278x9.
any new results yet? any higher clocks or fsb with the new bios?
Thats interesting... prime will die at 2 hours at 280 x 10 1.66v, I need like 1.69v to stabilize even then its not 100% stable - but I can run 312 x 9 1.64v and prime all night. Maybe thats just something with the chipset. I just figured it was my CPU being ghey.Quote:
Originally Posted by Atomicpineapple
I flashed with my pump off :D
OMFG !!! I just can boot at DF , oc will refuse booting , i flashed back 414-1 :(
I've done that once. After flashing, when I had reset all my figures, it gave me so many checksum errors. Wondered why, then pooed myself when I realised my pump was off. :DQuote:
Originally Posted by cpulloverclock
Hmmm, wonder if it is fixable in BIOS, be annoying if it isnt, oh well, could just be a few CPU's, not sure how widespread the problem is.Quote:
Originally Posted by dnottis
2772 1.41v ok
wc and room temp = 31C :D
cpulloverclock
Is that better or same as before
dunnoQuote:
Originally Posted by IvanAndreevich
will know in a few hours
Anyone know what mem table is in the 615 official bios?
just bought this mobo, ill give this bios a try later
i think 310Quote:
Originally Posted by bottone
Using 618-3, this one is better than all previous bios for me.
Onepagebook
You got some specific numbers? :)
yea, everyone says better but define better. I have never in all my life been able to get a higher fsb out of a bios than a previous bios. My system performs the same on 510-2 bios as it did on 1/25. No worse but certainly no better.
btw what is SLI broadcast aperture in the newer bioses?
looks good? :p:Quote:
Originally Posted by IvanAndreevich
http://img44.echo.cx/img44/2734/275gq1m9ab.jpg
http://img92.echo.cx/img92/3104/0513mpmw0mw.jpg
forgot to set trcd as 3 :p:
http://www.ocxtreme.org/onepagebook/.../3489pi32m.jpg
what makes it better? higher fsb/htt, higher stable cpu, runs cooler? what?Quote:
Originally Posted by Onepagebook
Of course that looks good ;) BUT what did it look on the previous bios is the question
OPB. Can you every ask Oskar if he plans on putting video card overclocking into the DFI Bioses? MSI and Gigabyte both have this option in their NF4 boards and I think it is a really great idea. People are having great results with it and it seems more stable than software versions in windows plus no additional software needed to be installed to overclock your video card if it can be done via the bios. I'm sure Oskar with his talent can pull this off and would be quite an addition to the DFI boards.
Quote:
Originally Posted by cantankerous
I asked that before NF4 series hit the retail, but I think Oskar has it's concern.
The hard part is there is no more space can put it in there with memtest86+ already there.
you're right except runs cooler. :)Quote:
Originally Posted by HousERaT
OK that's a little more clear. Do you have any numbers? Like what was the best before you used the 618-3 and what have you gotten since changing?Quote:
Originally Posted by Onepagebook
Hummm
Let's try...
I flashed back 618-3 and it works now , but the Vdimm measured by multimeter always over 0.1v than Vdimm in Bios setting ??? ( with all bios )
Flashed it. 3'2V+0'3Enabled gives me 3'08V and fluctuates to 3'18V (that didn't happended if I remember correctly). And it detects me 1023MB not 1024 now :?
Still at 236 1'5-2-2, 239 isn't stable like before (with jumper to 3'3V)
better?Quote:
Originally Posted by Onepagebook
and u posted that
I dont believe it's possible at 2800, especially with XP SP2 and these timings
28.234 it's the best time at 280 1:1 1.5-7-12-2-2-2 Win2k3
;)
http://www.ocxtreme.org/onepagebook/...bn/28gpi1m.jpg
others Japanese think the same thing :)
I'm waiting ur response impatiently :p:
:D Stayed a long time with 331-1. Working fine with BH-5 (max. 283 at 3,72 Vdimm) but when I got my X850XT I had a start up issues. BIOS recognized the card only when hitting the reset button.
Flashed to 615 and the issue disappeared, working fine and stable with my previous settings like shown in SIG.
Though I use PS/2 for mouse & KB i copied some pics from digi cam via USB and it worked fine for me.
:toast:
Hmmm ok, well thanks for looking into it. I am a little dissapointed I guess as I think this would be a really neat feature. I guess you can't have it all :rolleyes: . Hopefully something can be done in the future.Quote:
Originally Posted by Onepagebook
Did anyone notice that the 32byte Granularity setting went from (disabled 8 bursts) to (disabled 4 bursts) and from (enabled 4 bursts) to (enabled 2 bursts)?
Wonder if this affects performance...
Just flashed to this bios. No problem whatsoever yet no visible improvements yet either.
Works good with 2GB ram also :)
Any success on the cold boot problem with this BIOS?
I think these are the best BIOS yet. I am using 0618-2 with Gold BH and saw increased stability and OC. My previous best 32M run was 2815Mhz, now I can run 2840Mhz for the first time (I have run 3 32M so far). I can't wait to try out 0618-3 tomorrow with my TCC5. :D
Great work Oskar! :toast:
http://members.shaw.ca/Eldonko/258x11-1.JPG
I think most would prefer GPU OCing stuff over memtest in BIOS, I still use a floppy for memtest myself.Quote:
Originally Posted by Onepagebook
This is the best bios for me too.
This is not "a little" better than the other ones, this is VERY MUCH better. With other BIOSes, when I tried to pass memtest86 test5 with 2.5 3-3-6 I got around one or two thousands of errors, with this BIOS, I got less than a hundred.
Hope the next BIOSes can help me making my G.Skill's stable at 300 2.5 3-3 :D
Thank ya'
Everything WORKS!!! Yeahhhh!!! Finally. USB mouse works, usb keyboard works, ps2 works, and clocks are still great. Thanks Oscar. (618-3)
Seems like this is a killer bios, my cpu seems stable at 2,7ghz !!!
revived my 1GB of XP4000! :toast: Same overclock, will try to push farther later. Im just so happy my RAM isnt dead!
so,it appears that solves the cold boot issue! correct me if i am wrong, i almost receiving new sticks of redline, looking foward to test
618-2 working well for me, in fact my 250x11 Vdimm requirement seems to have dropped to 3.2V? Was 3.3V with Hellfire's 510-2.
Got alot of testing to do still, working on volt-crippled UTT-CH sticks now to see how they behave. Then Ill do new-BH ValueVX.
HARDCORE.... X850's are 5Vrail hungry.... I like to feed it 5.20V min.
dammit....had her running at 255x10 and 260x10 without errors...cranked it up to 265 and got some errors in memtest, shut it down, started back up and my cold boot is still here. :mad:
618-2 is looking good for my original BH-5 ram on 3000+ venice!
on 510-2
2x 512MB Kingston KHX3200/512 BH-5 ram in orange slots at 260mhz 2-2-2-6 needed 3.51v vdimm and Interleave disabled to even get pass test 1-3 error free, but still errored once in test #7
now on 618-2, 2x 512MB KHX3200/512 BH-5 same orange slots at 260mhz 2-2-2-6 did a full error free run at 3.41v vdimm with Interleave enabled :D
more testing of course and to think on P4 intel platform KHX3200/512 double sided ram was my worst BH-5 for P4 hehe :D
agreed.Quote:
Originally Posted by Eldonko
Kicka$$ to see a new revision already.....I'm gonna go slap on 618-3 and see how things go...can't wait now. :)
EDIT: Still have the ethernet issue with 618-3 same as I did with 615...tried manually inputting my MAC address but no difference. :(
Angry_Games explains the difference between BIOS revisions 1,2 and 3 over at dfi-street:
http://www.dfi-street.com/forum/show...1&page=6&pp=15
Angry_Games :
"-1 very tight DRAM drive strengths and really only for tight timings memory like bh-5
-2 best for all types, BH-5, UTT, TCCD (nothing is good for Hynix though). -2 is still the most popular and the most compatible that we have found
-3 supposedly for TCCD but -3 doesn't do TCCD for me, -2 does and I always recommend -2 as it has fixed the most user issues with bios out of any of them."
Tried both -2 and -3 respectively as recommended (-2 for BH-5 machine, -3 for TCCD machine using Kakaroto tweaks for TCCD). No noticable changes so far. Works fine except that for -3 with Venice I had to reset all the settings and redo it, sort of scary moment in the beginning.
The BIOS flashes/works fine for me...but I just can't get my GB lan working...argg...
Didn't seem to do much for my temp. corsair value select.
read the thread ! lolQuote:
Originally Posted by Cornelious0_0
Currently on 5/10-2 bios. Tried the 6/18-2 and -3 versions. No problems, but can't clock as high. Still testing though. There are some different ram settings in this bios. Might end up flashing back.
I got the same problem man, tried the orange dimm slots?Quote:
Originally Posted by 4X4N
Cold boot problem gone 618-2
thank you very much for this BIOS.
Peace Nugz
I just flashed my Bios to 615 yesterday. It solved my Cold Boot but decrease my CL timing for my TwinMos SP (from 2-2-2-2 to 2.5-2-2-2)in order to get it stable occasionally. I said occasionally becoz, sometimes, my windows would hang when ruuning application such as Prime 95, Sisofsandra, gaming (Far Cry) and etc. Did not encounter USB problems as I am using ps/2 mouse and keyboard. Will try to flash my bios to 618 in a few days. Anyway, good job Oskar becoz I am tired of 'living with cold boot'
well, she boots. Still have a bit of a cold boot issue, but shes been doing okay now. Wont do 260x10 stable anymore, but at least i can see Windows again :)
flashed to 618..................boot sequence definetely faster
I NEED my pc until Tuesday but I will see if my cold boot is gone on Wednesday :confused:
let's hope that Oscar cured cold boot with his changing the ram detect to after VDIMM is applied (from DFI street)
Quote:
Originally Posted by OSKAR_WU
occt 30 min, previous settings:
618-1: unstable (5 min pass)
618-3: unstable (5 min pass)
618-2: unstable (26 min pass)
(mbm) 3.46VDIMM -> 3.36VDIMM @ 247, unstable.
618-2:
(mbm) 1.56VID -> 1.54VID
stable (30 min pass)
ill try for higher when i get some time.
Sry, I guess the one thing that I never checked was the MII/Pin Strap to see what it was at...hopefully that's all it was, because I really wanna give 618 a try.Quote:
Originally Posted by Major
I'm gonna finish up my SuperPi 32M testing with this chip first to see where I top of for 32M stable and then see if i gain anything out of going to 618-2 as opposed to 510-2.
also, i still cant see my cdrom in bios under 'bootable', so i cant boot off of it (it shows up as a squiggley char.). issue started with 5/10. it is detected properly though by bios under dev. list/management.
Which CD-ROM you use ?Quote:
Originally Posted by pershoot
so is it safe to go back to the 5v for the vdimm using this bios?, atm i'm using the 3.3+ trick but it make my ram unstable @ 260 but runs fine using the 5v line.
I am wondering the same. Also wonder how the yellow slots are performing!?Quote:
Originally Posted by Wingz
No luck for me, still have the cold bug problem.
Been playing around with bios for a couple of days now. With my OCZ TCC5 the 5/10 bios seem better. I have some twinmoss bh-5 that had the cold boot problem with 5/10 bios that I thought I would try. So switched to the 5v line, got them up and running 255-2-2-2-6@3.5v. Not sure if it's completely stable but so far 3dmark 2001, 2m pass super-pi, sandra, pcmark all stable. Then turned off pc, unplug power from wall, wait 30 min. Plug back in, turn on, and boots up no problem!! Now understand even warm boot with these was not possible with 5/10 bios. I had just about given up running any high voltage ram. This is great :banana: Hopefully it will work for others too.
thanks mmw, i jest try the 618-3 and its work fine with my tccd, and it solve the usb mouse :clap:
I just got a x2, 4600 from the Egg. In my DFI with the 310P BIOS, while is runs quite nice, although it was about the same as my Venice 3200 when set to the same clock. Mem is A-Data 2X256 DDR600 Vitesta...
I flashed the 618-3 BIOS, resetup the BIOS, and now my SuperPI score is almost double!
I see 1 new RAM option called "DRAM Response Time". Tried all 3 settings and still SuperPI is SuperSlow....
Anyone got any idea why the 618-3 BIOS did this before I flash back to 310P????
aer you sure you re saved all your htt and multiplier settings to the correct ones? Something in your setting must be off.
Yeah, I had the whole mess written down... I could see being off a little, but flashing the 618-3 gave me 53 seconds SuperPI @ 9X300 @ 200 1T 2.5-3-3-7 with TCCD!
Can someone compare 510-2 hellfire bios with 6/18-2 from oscar with sisoftsandra (ram-cpu run these test twice) on the exact same settings with BH5?
I can't test it since I don't have a cooler atm for my cpu.
If someone is so kind to try this (will take 15-20min) because with hellfire's bios i had a big improvement on the memory speed versus other bioses.
Verified that this version does indeed set Vmem at initial cold-boot like it should :D :clap:
Does nothing for the "power applied to DIMMs" at cold-boot problem (but then, think only a MB revision will do that).
With TCCD and -3 BIOS, the "fastest" setting of "DRAM response time" limited OC to 30Mhz less than 331 BIOS. The "fast" setting reduced OC by ~5Mhz. The "normal" setting gave same results as "fast" setting. Note that this was with leaving all other settings exactly as they were with 331 BIOS.
Noticed system wouldn't initially boot fully (would hang at "Updating DMI") with any OC to speak of... had to get through one clean boot before any real OC'ing was possible (note: did a complete CMOS clear using command line switches of flash program when updating to 618).
it is my understanding that some settings, like the "normal/fast/fastest" only apply if certai nother settings are on AUTO. (Eg. max async latency etc.)
In other words:
If i use those settings (normal/fast/fastest) but i have my MANUAL values set in all other fields - do those settings anything ? (I see them only as abbreviations for which otherwise can be achieved by setting manual values elsewhere)
Cold boot issue gone here.. Thanks..........
Flash my bios from 615 to 618 last night. Tried all series. Stick to 618-2. My tweaking results:
618-1 - 260mhz@2-2-2-2@3.66vdimm-SuperPi 1m failed but running application and games fine
618-2 - 260Mhz@2-2-2-2@3.66vdimm - SuperPi 1 m failed but running applications and games fine
618-3 - 260mhz@2-2-2-2@3.66vdimm-Hang in windows
618-3 - 260Mhz@2.5-2-2-2@3.66vdimm - SuperPi 1 m failed but running applications and games fine.
For my case 618-3 behaves like 615 because I have to lower my TCL to 2.5 in order to load into windows
615 and 618 series is very picky on RAM timing. With Bios 331-1 I can run SuperPi 1m without fail at the same setting in bios 615 and 618. In short Bios 331-1 can push my oc much more stable than Bios 615 and 618 series. But too bad, I have to chose 618-2 cause I was having cold boot issues and can't go back to Bios331-1. So, nothing is perfect in this world. There is a gain and a lost. Anyway. I would like to thank Oskar in providing the Bioses and make the miraccle happen.
my sig
----------------------
Winnie3000+@2340Mhz(260 x 9)@1.52vcore(stock cool)
DFI NF4 Ultra-D (Rev. A3) Bios 618-2
TwimMos SP UTT 256mbx2 @ 2-2-2-2@3.66vdimm
IDE Hitachi 80gb 7200rom 2 mb
Gigabyte 6600GT 571/1140 (Stock cool)
LG 16XDVD ROM
Gigabyte 52XCDROM
Panasonic Floopy
icute 420watt PSU (20 pin)
Sandra sucks....it's only for show and not accurate, very inconsistant, if you run the same memory test couple times straight you will get different results every single time, same as everest. :stick: I prefer to measure with Superpi 8~32M.Quote:
Originally Posted by 187(V)URD@
By the way, for me 618-2 is slightly more stable and consistance than 510-2 or 414-2, prime95 can stretch much longer, didn't gain higher clock though. The "FASTEST" is not usesable and always hang on the DMI screen. Other than that I prefer this new bios over any other I've tried. I'd also noticed the vcore sensor is a bit more accurate now. :toast:
I keep getting the install new hardware thing from windows. It wants me to install nvraid or something like that.
Well go and disable whatever it wants you to install in the bios cause it was obviously disabled before and now after flashing and clearing CMOS it is once again enabled.
:( Same here - 331-1 was fine all day long - good runs in SuperPI but when I got my X850XT I had cold boot probs with the vid card - tried all catalyst versions, nothin' helped me out. Had always to push reset, then the monitor signal was given to the screen.Quote:
Originally Posted by longsiew
Now I flashed to 618-2 (tried also -3 & -1) and the issue is gone but no stable SuperPi aymore. Why?
Left all settings as in 331-1, DRAM resp time is set to fastest.
:toast:
I mostly run sandra 3 times and write it down, first run is always lower and the other 2 are always the same with me.Quote:
Originally Posted by ben805
But yeah superpi could be better :)
dram response time set to fastest is a no go for me. POST halts just before it says verifying dmi pool data it then reboots itself and does this over and over. I can only have my ram set at FAST. Perhaps fastest is very unstable?
i cleared CMOS (which I should have never done), and now i am unstable. i knew that was going to happen. im never clearing cmos ever again. i had stopped clearing it a while back (because of its inconsistency), and used the 3 consecutive flash routine (flash -> load optimised -> flash -> load optimised -> flash -> load optimised -> configure), as that was working well and was consistent. now i need to get out of this mess, hehe.
guess its time to flash back to 510-1 -> 618-2.
lesson to be learned here folks. if u are inconsistent with CCMOS after flash, dont clear CMOS unless you can't POST and have to!
Quote:
Originally Posted by pershoot
That's rather disturbing...... why would CMOS do that to your system unless maybe the flash was somehow unsuccessful?
my flashes are always successful houserat. its the chipset/board. same deal used to happen with my nf7 back in the nf2 days.Quote:
Originally Posted by HousERaT
computers. cant live with them, and cant live without them! hehe.
Hardcore, I feel your pain with the cold boot issue with video. If I use the 5v jumper, my 6800GT's in SLI will boot fine for teh 1st 2-3 times, but then after that, I get no video until in WinXP. Not even hitting the reset switch fixes it for me though. My solution was/is to get off the problomatic 5v jumper, do the VDIMM jumper mod where I'm running off the 3.3v line and tweaking the 3.3v rail. Its been totally rock solid now, and so far, lnock on wood, I haven't had problems with the video now posting. It seems that the cold boot issue is gone as well, but then I really don't unplug my system all that often.Quote:
Originally Posted by HARDCORECLOCKER
I really think that in the end, we are all going to find out that the cold boot issue is really going to be a board revision, and that a bios update isn't going to really fix it for us. Maybe thats 1 of the reasons the redesigned the boards. Maybe they fixed the regulatory ckt. or actually found a fix in another department? Who knows... I have a A8N Fati1ity SLI sitting here that I may crack open and try... So far, the cold boot issue hasn't plagued Abit yet, BUT they do seem to have a 1T problem at higher HTT speeds right now...
Dunno if it is mobo issue to be honest on 1T.... seems to be more chip related issue. But granted how new the board is... I haven't seen many winchester or... gasp... new castle users on this. Only winchester user I think was having no issue with 1T.... I still do have 1T issues on both my DFI and abit with venice.Quote:
Originally Posted by SVTSnake
618 seems to have an X2 issue....510 seems a LOT better with X2 so it looks like Oskar has some work to do still
You must keep in mind, this bios update has alot of changes in it besides the coldboot tweak.
I had significant improvements in Vdimm required/Mhz etc.
You need to look at 618-2 as a new bios and start from scratch on the timings.
Also, any BH based stuff wont like the "bank interleave" enabled from what Ive seen. but the impact on performance is minimal.... 26m10s vs 26m15s SuperPI result differences, etc.
So flashing to 681-X and expecting the same settings in CMos to get the job done is not sensible. Its back to the drawing board as if its a new board and you need to find the max OC all over again.....but I think you'll ultimately be rewarded with a higher max OC.
But as an example, my 4x512M UTT-CH setup dropped from requiring 3.3Vdimm to now only needing 3.2Vdimm to achive the same 100% stable 250Mhzx11 results.
And my sets of TwinmosSP with voltage problems (they lock up over 3.2Vdimm) work to 250Mhzx11 now without pushing the Vdimm to the breaking point.
618-2 has brought me nothing better or worse than 510-2. I can use exact same timings, clocks etc and get the same stability, scores etc. Can't get any higher but also not lower. Bank Interleaved enabled works just fine on my 1gb BH6 chips too.
i wish i could turn off bank interleave, hehe. turning it off creates thousands of errors on my SD core. the winnie had no issue with it though, however. its very interesting to see similar parts, act so different in their respective environments. such is life:)Quote:
Originally Posted by uwackme