Soyo P4S Dragon Ultra problems

I run with a 1,8 Wil. and can only get to 112 fsb then it wont post anymore
(bios 2BAA1

Have just ordered a NW 2.2 and wil try to see if that changes anything

Have anyone tried to OC the NW 2.2 - and how far will it go on a P4S?

"Steve" <Somewhere@overtherainbow ...

I have not done any real benchmarks on this set-up yet. I am still trying
to determine how fast I can push it without hosing Windows. I'm still at
the "boot to DOS with HDDs disconnected" stage. That's why I'm looking for
a DOS Util that can read the speed of the PCI bus. Soyo have been remiss in
not showing this information in the BIOS screen. I need to work out at what
FSB the board switches from a 1/3 to 1/4 PCI multiplier. I know my AGP card
is good to 100 MHz, but the Highpoint and SiS IDE can be finicky with PCI
speeds, especially when DMA transfers are involved.

However, I do have some Sandra 2002 (2002.1.8.59) memory scores from my old
1.6 Willamette with some CAS2.5 PC-2100 Crucial DDR (Win98 SE). The OCZ
will give similar scores if the same settings are used:

100/133 MHz FSB/DRAM
Normal Timing = 1825/1818 (ALU/FPU)
Fast Timing = 1859/1829 (ALU/FPU)
Turbo Timing = 1855/1836 (ALU/FPU)

100/167 MHz FSB/DRAM
Normal Timing = 2116/2093 (ALU/FPU)

105/140 MHz FSB/DRAM
Normal Timing = 1919/1900 (ALU/FPU)
Fast Timing = 1964/1936 (ALU/FPU)
Turbo Timing = 1979/1945 (ALU/FPU)

105/175 MHz FSB/DRAM
Normal Timing = 2232/2206 (ALU/FPU)
Fast Timing = 2254/2221 (ALU/FPU) - some video corruption in Windows

There was plenty of stuff running in the background and these are single-run
numbers that I happen to have on me, so these are only ball-park figures.
Yours should be slightly better if you are running without the background
apps.

In any case, Sandra is not a good benchmark tool. I have done 1 run of
MadOnion's PCMark 2002 (#), but
that's about it. When I find the "sweet spot", I'll be running proper
WinBench @ Quake tests.

The Crucial starts losing data integrity (in Memtest86) at about 139 MHz
(Turbo Mode), 182 MHz (Fast Mode) and 183 MHz (Normal Mode). The OCZ starts
to lose it at about 193 MHz (Fast Mode) and 205 MHz (Normal Mode). I have
not had time to test the OCZ in Turbo Mode yet, but so fat it is good to 176
MHz.

So far I am well impressed with both the OCZ memory and the Northwood P4. I
may have a different opinion when I move onto the Windows tests, but that's
a story for another day...

Al.
--
"Instructions are only for people who don't know what they're doing"
Bob the Builder - Can we fix it? (2001)
--

All messages sent to me will pass through Spamcop.
If you Spam me, you will end up on a black list and
your ISP will receive an automatic complaint report.
In any case the Spam will not be read before it is trashed.
If you STILL want to Spam this address, here are
more people that would love to hear from you:




--
...

Why wait??? Take a 1.6A or a 1.8A and clock them at 133mhz. That
will have them running at the 533mhz. The 1.6A is now a 2.133ghz
533mhz FSB and the 1.8A is now 2.4ghz 533mhz FSB.

The northwood processor is very overclockable. I have two of the
1.6A's running stable at the 2.133ghz at default vcore on a 645 MSI
Ultra (6547) motherboard with samsung PC2700 DDR.

Thus I gave a

P-IV Northwood at 2.133ghz, 533 quad pumped FSB (133mhz clock)
DDR is PC2700 clocked at 333mhz (Memory is 4:5 ratio).
Agp clock is 66.6mhz and PCI clock is 33.3mhz (Fully compliant)

These systems are extremely powerful.

Steve

thats another... im stuck to deciede if i should upgrade to a NW or
stick to my 1.9Wil i kinda want to wait til the (533) bus P4s
comeout... btw Al... can u post ur OCZ mem bechs I have the 2700 rev
1.. (which i might replace soon) and i wanted to know if my benchs were
ne where near what they should bethankx

Hi Will,

The 2BA2 BIOS does not seem to work well with Willamette P4s. I had a 1.6
that would overclock to about 115 MHz FSB using BIOS 2AA2. Windows was
reasonably stable at up to 112 MHz FSB, but it locks up after 24-48 hours at
that speed. I flashed to 2BA2 and it would not boot at anything other than
100 MHz FSB. A CMOS reset was all that I could do to get it started after I
set the FSB to 102 MHz. Others have had similar problems. See the thread
entitled "new p4s bios... ne one use it yet???" for more details.

Last night I upgraded to a 1.8A Northwood and 256 MB OCZ PC-3000 rev 2.
Using a modified version of the 2AA3 BIOS
(#) this boots and runs
Memtest86 cleanly at up to 138 MHz FSB (not had time to test Windows
stability). However, if I set the DRAM clock to anything above the FSB
speed when the FSB is set to 133 MHz or above, then it fails to boot.

It will work with 132/176 MHz FSB/DRAM (Turbo Mode) or 123/205 MHz FSB/DRAM
(Normal Mode), but 133/166 MHz FSB/DRAM fails in any mode.

Someone going under the name of "me" has posted that the 2BA2 works well for
him (he is using a Northwood). My (as yet unproven) theory is that the 2BA2
works great for Northwood P4s, but has a problem with Willamettes. If I
were still using a Willamette, I would definitely stick with the 2AA3
version.

Now that I have a shiny new 1.8A, I am going to try the 2BA2 BIOS some time
this week to see if it fixes the 133 MHz FSB/DRAM problem. I'll post the
result when I'm done testing.

Al.
--
"Instructions are only for people who don't know what they're doing"
Bob the Builder - Can we fix it? (2001)
--

All messages sent to me will pass through Spamcop.
If you Spam me, you will end up on a black list and
your ISP will receive an automatic complaint report.
In any case the Spam will not be read before it is trashed.
If you STILL want to Spam this address, here are
more people that would love to hear from you:




--
...