v mod 3/4 for p4b533

[dcranford]

Ok,

Here is my info regarding the P4B533-E, I have a 2.26 Northwood B, 512 MB Kinston 2700

I have been able to get stable in windows XP at 150 FSB with this new trick before using 1:1, I had been able to boot into
166 FSB. IS it worth Buying better memory to run at 166FSB with this new 3:4 mode or will I get this same performance at 150?

 

[THUGSROOK]

 

[THUGSROOK]

 

[NanoMem]

Originally posted by: THUGSROOK
NanoMem ~ you are confused - we are in jumper free mode this time
all you have to do enable the correct dipswitch on your board to get the 3:4 (or 4:5) ratio in jumper free mode. there are ZERO limitations this time - all bios functions work perfectly.

I know that this trick is being achieved with Jumper Free Mode i.e. jumper "JEN" left at its factory default setting of pins 1-2 jumpered! You're just misunderstanding my statement/question. What I was trying to get from you is the actual cpu:ram ratio response for those two DSW1 settings in Jumper Free mode...

Again if you go back to my post above you will notice that even though the last switch 6 on DSW1 is activated the frequency readings off DSW1 are actually less than 133MHZ. In fact they should be 120 and 125MHZ (make sure to scroll down to see the table of values) respectively which puts them below the 133 MHZ line even though switch 6 is in use.

Switch 6 is not exactly "undocumented" as it serves to set values for the external CPU bus frequencies when Jumper Mode is enabled which also causes the BIOS cpu clock values to be ignored at the same time. But it appears based on this thread that in Jumper Free Mode, this switch is having an impact on overclocking RAM through an implicit ratio increase while the CPU is still receiving its external bus clock values from the BIOS. RAM is also still receiving the same BIOS clock settings as the CPU i.e. from the BIOS but with a different ratio when switch 6 is flipped around.

I have observed that on the P4B533 line of boards that in general the purpose of this last switch, whether number 5 (P4B533V, -VM or -M) or number 6 (P4B533 or -E) is to cross above 133 system bus frequencies. If you take the time to sift through the DSW1 tables of all these boards you will note that P4B533 & P4B533-E have the two exceptions to this argument as I am still inquiring about them from my first post above.

Answers to my problem should shed some light in determining whether this is a hardware bug in the RAM bus lines such as a possible limitation in the external clock generator, its trace lines to RAM or the RAM clock multiplier that Asus can't circumvent. In this case we can rest assured that it won't be fixed and we can happily live with it as long we own the boards. Otherwise, this might also be a simple BIOS code bug where the BIOS cpu:ram ratio is ignored if the last switch is activated and in this scenario, Asus could easily rectify it with a future release, taking away the benefits of such "nice bug".

Btw, I am also curious if this works at any BIOS version for all the boards. Might consider adding BIOS level info indicating when trick works or not.


Peace

 

[THUGSROOK]

hehe yes i was a little confused

but all it really comes down to is that DS6 (on p4b533 boards) is a 3:4 ratio enabler weather in jumper free mode or jumper mode. (i dont think anyone wants to bother with jumper mode tho if they can avoid it - so lets just talk jumper free mode)

(those jumper charts is what gave me the idea in the 1st place)
i also noticed that DS5 is always opposite of DS6....
could there be more hidden features in there?
is there a combo for 4:5? or 5:6?

this project aint over yet - not by a longshot

as for bios revisions - so far its working for everyone that has good enough ram. if there are any bios revison problems i will keep notes and post all available info. (as always)

 

[oldfart]

It will be interesting to see Duvie's video encoding benches. I do a bit of that myself. I think he does DVD to DiVX ripping. I've been doing some AVI capture to VCD/SVCD (MPEG1/MPEG2) encoding. If I get a chance, maybe I'll bench that next time I do one.

 

[NanoMem]

Originally posted by: SupermanCK
Wait...I am a bit confused now...
so is it dipsw 10 or 9???

SupermanCK, It's switch 9! The reason is if you scroll up and down the table of undocumented values that Thugs established at the time, you'll realize that 9 is the switch that causes bus frequencies to cross above 133 MHZ. Furthermore 3:4 is already achieved (although differently as jumper mode is required for that trick to work)
I don't have the P4B266 board but if you really want to apply the trick from this thread, you first need to re-instate "Jumper Free Mode". Next make sure that when you go above 132 MHZ in the BIOS, the 3:4 ratio doesn't work i.e. CPU:RAM=1:1. After that flip around switch 9 and see if you get the same results as the P4B533 i.e. RAM implicitly clocking at 3:4 above 132MHZ. If switch set to ON or OFF doesn't work then as a last resort try switch 10 which I strongly believe will not work but you can always try.

Bottom line you need to test the following 4 settings for 9 and 10 (don't worry about settings for switches 1 through 8) and remember all this under ?Jumper Free Mode? and BIOS frequencies above 132.

Switch 9, Swicth 10
ON, ON
ON, OFF
OFF, ON
OFF, OFF

Keep us posted!

Peace

 

[SupermanCK]

alright...i got it...i will try these setting later on tonight...thanks thugsrook and nanomem

 

[NanoMem]

Originally posted by: THUGSROOK
hehe yes i was a little confused

but all it really comes down to is that DS6 (on p4b533 boards) is a 3:4 ratio enabler weather in jumper free mode or jumper mode. (i dont think anyone wants to bother with jumper mode tho if they can avoid it - so lets just talk jumper free mode)

(those jumper charts is what gave me the idea in the 1st place)
i also noticed that DS5 is always opposite of DS6....
could there be more hidden features in there?
is there a combo for 4:5? or 5:6?

this project aint over yet - not by a longshot

as for bios revisions - so far its working for everyone that has good enough ram. if there are any bios revison problems i will keep notes and post all available info. (as always)

Remember my question is only about two settings and in "Jumper Free Mode" of course. But if I you're curious about sniffing out 4:5 or 5:6, I have 30 more combinations in additions to my two. Let me know if you want me to post all of them for you.

Peace

 

[THUGSROOK]

 

[THUGSROOK]

NanoMem ~ YHPM

 

[SupermanCK]

i have tried both settings...still no go...

 

[NanoMem]

Originally posted by: SupermanCK
i have tried both settings...still no go...

Well, it looks like there is no new history to be written for the P4B266 although 3:4 is achievable in "Jumper Mode". Only inconvenience is its loss if CMOS is ever accessed again after 3:4 has been activated. Sorry to hear it didn't work out but it definitely was worth a shot!

Peace

 

[Sharpio]

Can anyone confirm this works with the P4B533-V too??

 

[Bobbyeyes]

the v has this great stuff already

 

[Bobbyeyes]

UPDATE ON MY LITTLE P4B533-VM
it has both 4:5
and 3:4
(WHERE 1:1 IN THE BIOS MEANS 3:4 )

i can 0NLY achieve this by TWO THINGS:
using the ORIGINAL bios (whitch is no longer avail. on the asus web site)
the ratios are all there up until 170fsb

and our magic jumper

the "new" bios,and both beta versions 004-asus site, 005-german site
are even worse than the final new bios

Thugs,
is there any way you can try to get a hold of the original bios on your 533?
i think you posted once-any luck?
i hope someone knows of a way to get thugs the original bios.some one who has the board, never upgraded, and can put it on a floppy and send it to him

my point is:
the above "features" dissapear on the upgrade bios versions
this could only be unique to my board...but i think it could apply to all the 533's

 

[Bobbyeyes]

afterthought

that also explains why i could never boot @141,143,145,150-170mhz
when the mem ratio was set to 1:1
it really meant 3:4
i could only boot @ mem set to auto
which means 1:1

 

[THUGSROOK]

 

[SupermanCK]

thugsrook if you are not happy with your 3:4...then i should be very unhappy with my manual 3:4...after the hours i spent already trying to find a way to get jumperless 3:4...maybe i will try some more option tonight to make sure there isn't any

 

[THUGSROOK]

whaaaaa! i want a 5:6
(166fsb 400ddr)

 

[THUGSROOK]

Bobbyeyes ~ im little confused as to how your VM board now works....

1:1 = 3:4
?:? = 4:5 ?
Auto = Auto?

 

[Bobbyeyes]

in the bios:
auto = auto (means 1:1)
1:1 = 3:4
5:4 = 5:4
4:5 = 4:5

Thease are the four choices that ONLY appear on the first bios
these choices dissapear at 180fsb and above.i believe
there is only auto and 5:4 above 180 but 180 does not pertain to me.

another thing about the vm ... i have usb and the onboard lan card enabled, and they work.
does not seem to effect the overclocking of the cpu....
w/ the 266-c, i could ONLY achieve a 150 or above fsb w/ the usb disabled

 

[THUGSROOK]

what bios version is that?

can you email it to me? (my profile) ill post it for all to get.

 

[Bobbyeyes]

give me till tomorrow...
i will do it @ my office
it is the first version which is no longer on the asus web site

 

[NanoMem]

Originally posted by: Bobbyeyes
UPDATE ON MY LITTLE P4B533-VM
it has both 4:5
and 3:4
(WHERE 1:1 IN THE BIOS MEANS 3:4 )

i can 0NLY achieve this by TWO THINGS:
using the ORIGINAL bios (whitch is no longer avail. on the asus web site)
the ratios are all there up until 170fsb

and our magic jumper

the "new" bios,and both beta versions 004-asus site, 005-german site
are even worse than the final new bios

Thugs,
is there any way you can try to get a hold of the original bios on your 533?
i think you posted once-any luck?
i hope someone knows of a way to get thugs the original bios.some one who has the board, never upgraded, and can put it on a floppy and send it to him

my point is:
the above "features" dissapear on the upgrade bios versions
this could only be unique to my board...but i think it could apply to all the 533's

Hey Bobbyeyes!

Just when I thought this was a closed matter, you re-ignited it! I consider your findings crucial in terms of circumstances under which this trick would or wouldn't work.
Asus couldn't have been that careless in allowing a hardware bug or "undocumented" feature if you will, to slip through on such high quality boards. This feature was most likely left open by BIOS coders for the intent of engineering testing and verification of high FSB's with a 3:4 ratio at Asus. It is much easier to rectify a soft lock than a hard one so I would not be surprised if Asus decided to turn it off from now on with future BIOS releases. I'd definitely probe these BIOS dependencies as I suggested in an earlier post and determine which level closes the door.

btw, a 5:4 underclocks your RAM by 20%*2 (for DDR) versus CPU!

Thanks for the info.

Peace