v mod 3/4 for p4b533

[THUGSROOK]

 

[Mikki]

Thugsrook, I can't even imagine what you had to go thru. I don't think this problem is as big of a deal as I originally thought. Like you said, it's not a concern for everyone. I know I'll never be running @ 133fsb except to test stuff.

 

[THUGSROOK]

anyone hear from mlaine about his problem yet?

Mikki ~ did you run your tests with your USB enabled?

 

[Mikki]

Thugsrook, yep, and I was religious about doing it the same way for every test. The only variable that I couldn't do anything about was the v-core, but I did use the same voltage for each FSB speed, whether it was DS6 on or off. If you wanna know exactly what i did, lemme know.

 

[THUGSROOK]

i was just wondering if your USB ADSL modem is still causing a performance drop?
or did you resolve that?

 

[Mikki]

Yeah, I've resolved it: if I wanna boost I disconnect my modem, problem solved...
That'll have to do for now until I can afford an asdl router.

 

[THUGSROOK]

 

[SupermanCK]

Originally posted by: Mikki
Thugsrook, here's my info:
P4B533-E
256 Corsair XMS3200
WinXP

SciSoft Sandra - CAS . 2/2/2/6 . . . . . . 2/3/3/6 . . . . . . . 2.5/3/3/7
100fsb 3:4 266ddr = 1946 / 1945 . . . 1902 / 1903 . . . 1900 / 1900 DS6 ON
113fsb 3:4 300ddr = 2188 / 2189 . . . 2142 / 2143 . . . 2137 / 2136 DS6 ON
166fsb 1:1 333ddr = sorry, my cpu won't go that fast
125fsb 3:4 333ddr = 2434 / 2434 . . . 2381 / 2380 . . . 2379 / 2380 DS6 ON
133fsb 3:4 354ddr = * FAILED * . . . . 2577 / 2578 . . . 2573 / 2573 DS6 ON
150fsb 3:4 400ddr = * FAILED * . . . . 2882 / 2882 . . . 2878 / 2877 DS6 ON

100fsb 3:4 266ddr = 1946 / 1945 . . . 1903 / 1901 . . . 1901 / 1901 DS6 OFF
133fsb 1:1 266ddr = 1939 / 1941 . . . 1891 / 1890 . . . 1890 / 1889 DS6 OFF
113fsb 3:4 300ddr = 2190 / 2189 . . . 2141 / 2140 . . . 2137 / 2137 DS6 OFF
150fsb 1:1 300ddr = 2172 / 2172 . . . 2118 / 2118 . . . 2118 / 2117 DS6 OFF
125fsb 3:4 333ddr = 2433 / 2432 . . . 2380 / 2382 . . . 2379 / 2380 DS6 OFF

hey mikki...your memory scores seem a little low to both thugsrook and me...are you running some other program while benchmarking?
if you compare yours to thugsrook...then you will know what i mean...

 

[Mikki]

SupermanCK, check out my memory thread for a reply on that. And yeah, I've compared mine to Thugsrook's and...I...don't...really...wanna...talk about it....

HEHE

 

[NanoMem]

Originally posted by: Ardy
Well I tried for the 3:4 on the P4B533-V and if the correct settings are off,on,on,on,on and on,off,off,off for the 1.6a chip,then it's a no go. Hmmmm.... this is with the 1004.7 beta bios....I still can try it with 3 more bioosis ?

Ardy, are you by any chance not mistaking DS1 with DS5 on the DSW1 DIP switch box of the P4B533-V. DS5 is the one closest to the CPU. On your P4B533-V, only DS5 should have been flipped around without touching or changing switches 1 through 4. FYI, your OFF ON ON ON ON corresponds to 100MHZ if "Jumper Mode" was enabled. I could not make your next DSW1 setting from above as you're short one switch.

Anyhow, if you don't remember what your factory settings for DSW1 were, try the following
DS1, DS2, DS3, DS4, DS5
OFF, ON, ON, ON, OFF or
ON, ON, OFF, ON, OFF
all in "Jumper Free" mode by leaving jumper "JEN1" set at its factory default setting of pin 2 and 3 jumpered.

If 4:5 still can't translate to 3:4 above 132MHZ then you might be running into the same BIOS problem that Bobbyeyes pointed out with his board (his is the P4B533-VM). In that case you probably might consider downgrading your BIOS as you thought to one of the first May BIOS versions which I believe are still on their web site.

Oh, one more thing. Just out of curiosity, are you using the on-board video or a separate card in your AGP slot?

Keep us posted!

Peace

 

[THUGSROOK]

 

[NanoMem]

Originally posted by: Mikki
Thugsrook, here's my info:
P4B533-E
256 Corsair XMS3200
WinXP

SciSoft Sandra - CAS . 2/2/2/6 . . . . . . 2/3/3/6 . . . . . . . 2.5/3/3/7
100fsb 3:4 266ddr = 1946 / 1945 . . . 1902 / 1903 . . . 1900 / 1900 DS6 ON
113fsb 3:4 300ddr = 2188 / 2189 . . . 2142 / 2143 . . . 2137 / 2136 DS6 ON
166fsb 1:1 333ddr = sorry, my cpu won't go that fast
125fsb 3:4 333ddr = 2434 / 2434 . . . 2381 / 2380 . . . 2379 / 2380 DS6 ON
133fsb 3:4 354ddr = * FAILED * . . . . 2577 / 2578 . . . 2573 / 2573 DS6 ON
150fsb 3:4 400ddr = * FAILED * . . . . 2882 / 2882 . . . 2878 / 2877 DS6 ON

100fsb 3:4 266ddr = 1946 / 1945 . . . 1903 / 1901 . . . 1901 / 1901 DS6 OFF
133fsb 1:1 266ddr = 1939 / 1941 . . . 1891 / 1890 . . . 1890 / 1889 DS6 OFF
113fsb 3:4 300ddr = 2190 / 2189 . . . 2141 / 2140 . . . 2137 / 2137 DS6 OFF
150fsb 1:1 300ddr = 2172 / 2172 . . . 2118 / 2118 . . . 2118 / 2117 DS6 OFF
125fsb 3:4 333ddr = 2433 / 2432 . . . 2380 / 2382 . . . 2379 / 2380 DS6 OFF

Looks like most of my suspicions are getting confirmed! The trick seems to have a correlation with 133MHZ FSB's and above being hard-coded in DSW1 while still in "JUMPER FREE MODE" more than the mere flipping of the last switch itself. It's as if settings of DSW1 at or above 133 provoke clocks transferred to the RAM to be equal to whatever CPU FSB value is entered in the CMOS multiplied by the next higher ratio than the one used in the CMOS.

I wonder if Ardy is running into the same issue with his P4B533-V.

I originally made this observation just by figuring out the real purpose of this last switch on these boards. Read one of my posts above to see what I am talking about. Therefore I don't believe there is a problem with P4B533-E boards but just our initial perception of how this trick really works.

Great info Mikki!

Peace

 

[NanoMem]

Originally posted by: SupermanCK
that's weird...both thugsrook and mikki can't finish running mem @150fsb...
do you think there are some thing going on with this trick that is hurting your mem???

I think the issue has only to do with the most aggressive memory timings! It would be nice to ask them what ratings their RAM chips have if of course those aren't hidden by heat spreaders.

Peace

 

[Mikki]

All right, check THIS out:

SciSoft Sandra - CAS . 2/2/2/6 . . . . . . 2/3/3/6 . . . . . . . 2.5/3/3/7
100fsb 3:4 266ddr = 1946 / 1945 . . . 1902 / 1903 . . . 1900 / 1900 DS6 ON
113fsb 3:4 300ddr = 2188 / 2189 . . . 2142 / 2143 . . . 2137 / 2136 DS6 ON
166fsb 1:1 333ddr = sorry, my cpu won't go that fast
125fsb 3:4 333ddr = 2434 / 2434 . . . 2381 / 2380 . . . 2379 / 2380 DS6 ON
133fsb 3:4 354ddr = * FAILED * . . . . 2577 / 2578 . . . 2573 / 2573 DS6 ON
150fsb 3:4 400ddr = * FAILED * . . . . 2882 / 2882 . . . 2878 / 2877 DS6 ON

100fsb 3:4 266ddr = 1946 / 1945 . . . 1903 / 1901 . . . 1901 / 1901 DS6 OFF
133fsb 1:1 266ddr = 1939 / 1941 . . . 1891 / 1890 . . . 1890 / 1889 DS6 OFF
113fsb 3:4 300ddr = 2190 / 2189 . . . 2141 / 2140 . . . 2137 / 2137 DS6 OFF
150fsb 1:1 300ddr = 2172 / 2172 . . . 2118 / 2118 . . . 2118 / 2117 DS6 OFF
125fsb 3:4 333ddr = 2433 / 2432 . . . 2380 / 2382 . . . 2379 / 2380 DS6 OFF

After I disabled all USB, LAN, and Sound:
113fsb 3:4 300ddr = 2317 / 2317 . . . 2294 / 2293 . . . 2291 / 2290 DS6 ON
125fsb 3:4 333ddr = 2564 / 2564 . . . 2538 / 2539 . . . 2536 / 2536 DS6 ON
133fsb 3:4 354ddr = * FAILED * . . . . 2736 / 2735 . . . 2733 / 2732 DS6 ON
150fsb 3:4 400ddr = * FAILED * . . . . 3050 / 3050 . . . 3048 / 3048 DS6 ON

Those scores are much better, and from what I've seen, right on par for a -E board with XMS3200.
Thugsrook, one thing I've been meaning to ask is your memory scores generally have a large split between the first set and the second set, with the first being higher, whereas mine are pretty much even. Do you know why?

 

[Ardy]

Nanomem
I worked a little with Thugsrook on this last night and have been on it for the past hour.According to the original post of setting dips on DSW1 to 100 fsb (only 5 switches) I set to (with 1 being the first and 5 being the last) to off,on,on,on,on and DSW2 to 16x or on,off,off,off.Since you don't seem to be concerned with DSW2 I stopped messing with that and concentrated on the #5 switch of DSW1. The manual shows it on, at all Mhz below 133 and off at 133 and above.I have tried the #5 switch off and on in different switch positions (of 1-4) and with all memory ratios available.This is all done in jumper free I found no changes in Sandra scores from what I had before the changes.I did have a no boot (CPU not found) at Thugs suggestion of #1 on the rest off.(manual 166mhz setting) I tried A bunch of settings with no change so far.Don't know if I tried the ones you recommended but I will now.Thanks a bunch from everybody

 

[Ardy]

Also I want to add, this is all with the 1004.7 beta bios. I will try version 1002 if i can't get this to work.

 

[Ardy]

OK this is what I got.Odd your first set of settings (133mhz factory) was my last test.On to your other settings of on,on,off,on,off or what the manual shows to be 135fsb.

Scores were taken at 135fsb in bios.

Sandra memory scores:

4:3=1484/1512
1:1=2040/2063
4:5=2567/2601
Auto=2565/2597

Same scores as I've always gotten except I'm not sure about auto.Seems like it used to default to 1:1 but I'm not sure,rarely used it.

 

[THUGSROOK]

 

[NanoMem]

Originally posted by: Ardy
Nanomem
I worked a little with Thugsrook on this last night and have been on it for the past hour.According to the original post of setting dips on DSW1 to 100 fsb (only 5 switches) I set to (with 1 being the first and 5 being the last) to off,on,on,on,on and DSW2 to 16x or on,off,off,off.Since you don't seem to be concerned with DSW2 I stopped messing with that and concentrated on the #5 switch of DSW1. The manual shows it on, at all Mhz below 133 and off at 133 and above.I have tried the #5 switch off and on in different switch positions (of 1-4) and with all memory ratios available.This is all done in jumper free I found no changes in Sandra scores from what I had before the changes.I did have a no boot (CPU not found) at Thugs suggestion of #1 on the rest off.(manual 166mhz setting) I tried A bunch of settings with no change so far.Don't know if I tried the ones you recommended but I will now.Thanks a bunch from everybody

That's right! DSW2 has absolutely no effect whatsoever. The DSW2 dipswitch box is only intended as a multiplier to externally clock the core of "Unlocked"Intel CPU's. If you happen to be in possession of such a "gem" consider yourself very fortunate as your clocks would be setting new records as we speak. I haven't seen or heard of an Intel chip with unlocked core since the days of "BX" chipsets and that's years back, unless you happen to be an Intel chip insider.
Peace

 

[mlaine]

Hi!

I started another thread about memory performance on my P4B533-E but was kindly pointed to this thread by Thugsrook (thanks!). Here's a quote of my original message:

<START QUOTE>
Hi!

I'm having trouble getting satisfactory memory performance with my setup:

Intel Pentium 4 2.26 GHz/533FSB, Asus P4B533-E (BIOS 1004), Intel stock cooler, 2x256 MB Kingmax DDR400 DIMM, Enermax EGP-465VE 431W, PNY Verto GF4 Ti4600, SB Audigy Platinum EX, Intel PRO/100 S NIC.

I'm using the "dipswitch 6 trick" to get 3:4 FSB/memory ratio on my mobo. However when I activate the 3:4 ratio (1:1 setting in BIOS) the system is really unstable even with default 133 MHz FSB (can't even boot to Windows XP). The memory is supposed to do 200 MHz @ CAS 2.5-3-3 so the memory speed of 177 MHz yielded by 133 MHz FSB should be no problem. I've tried to run memory at default voltage and up to 2.9V and relaxed CAS 2.5-3-3-7 settings but nothing seems to help.

The processor is now running at 160 FSB (2.720 GHz) with 1:1 ratio and CAS 2-3-3-5 (SPD setting). I'm quite satisified with this but would rather run the system with lower FSB and greater memory speed. Is the memory crap or is there something else that could be holding it back? I do not know any resellers here in Finland with Crucial/Corsair/Mushkin memory and the Kingmax was the only non-generic memory brand I found above PC-2100.
<END QUOTE>

Now there seems to be a lot of info about this 3:4 ratio setting... I'm having trouble understanding which is related to P4B533-V and P4B533-E and how exactly this info relates to my P4B533-E situation? The dipswitches from 1 to 5 are all off and 6 is on. My USB controllers and RAID controller are enabled (can't turn them off as my bootable HD' is in RAID controller and mouse&keyboard in USB).

I upgraded to beta BIOS 1007-12 and locked my PCI/AGP through manual BIOS setting to 33.33/66.66 but it didn't help and the system is still unstable with 3:4 ratio active.

Any suggestions are appreciated!

 

[THUGSROOK]

 

[NanoMem]

Originally posted by: Ardy
OK this is what I got.Odd your first set of settings (133mhz factory) was my last test.On to your other settings of on,on,off,on,off or what the manual shows to be 135fsb.

Scores were taken at 135fsb in bios.

Sandra memory scores:

4:3=1484/1512
1:1=2040/2063
4:5=2567/2601
Auto=2565/2597

Same scores as I've always gotten except I'm not sure about auto.Seems like it used to default to 1:1 but I'm not sure,rarely used it.

I think you meant to say Auto has same results as 4:5. Your only alternative at this stage is to revert to an older BIOS rev. and see if this trick will produce 3:4 above 132MHZ FSB's. I have a P4B533-V but because I am using the on-board video my sandra readings are totally meaningless as video has to requisition a portion of system RAM. My Sandra RAM bandwidth gets totally butchered in the process as I can't even tell what my DDR is although I do know that 4:5 is in effect above 132 FSB's. Btw, what's your AGP?
Peace

 

[Ardy]

"I think you meant to say Auto has same results as 4:5. Your only alternative at this stage is to revert to an older BIOS rev. and see if this trick will produce 3:4 above 132MHZ FSB's. I have a P4B533-V but because I am using the on-board video my sandra readings are totally meaningless as video has to requisition a portion of system RAM. My Sandra RAM bandwidth gets totally butchered in the process as I can't even tell what my DDR is although I do know that 4:5 is in effect above 132 FSB's. Btw, what's your AGP?
Peace"

Yep,my Auto scores are the same as 4:5 but I'm not sure they were before.I thought they defaulted to 1:1 but don't hold me to that.Unless it's of interest I won't check.I'm running a Geforce 3 TI-200 on the AGP.

The reason I was putzing with the DSW2 was from the original post.

Procedure:

1) Power down (obviously) and set the DIP switches to 100MHz and whatever multiplier corresponds to your processor. Do NOT change the jumper to jumpered mode - leave it on jumperfree.

2) Power up and select 70/35 for the AGP/PCI frequency, and 1:1 for the CPU:Memory ratio. Select any FSB above 132MHz and an appropriate memory timing, and the actual ratio will be 3:4 at the timing you selected.

Setting the memory to 4:3 will cause the memory to run at 1:1; setting it to 1:1 will make it run at 3:4. Strange, but it works. Attached are a couple of Sandra screenshots.

We must have discarded most of his info such as the 70/35 thing,the multiplier ect...

 

[NanoMem]

4:5 ratio above 132MHZ FSB's has always been there on my P4B533-V as far as I can remember...
Peace

 

[THUGSROOK]