I have been holding-off on upgrading since the P4 640 @ 4.4GHz rig that I had was hard to beat IMO, but after seeing some of the Presler results roll in... I just had to get in the game.... 
I picked-up a retail P-D 920 and paired it with a refurb ASUS P5WD2-Premium mobo to lessen the $$ outlay.
The Processor specs:
Product Code: BX80553920SL94S
MM#: 878915
FPO/Batch#: L541B403
Pack Date: 12/30/05
I am using watercooling, which consists of:
Swiftech APOGEE waterblock
Swiftech Bay Res /MCP-350 pump
Swiftech RadBox external radiator mount
Heater core radiator
120MM Vantec Stealth fan with shroud
Other specs in sig....
I put the rig together, and went straight for 300MHz FSB... (4.2GHz) It stalled at the WinXP boot screen. :frown: I checked the NB heatsink on the ASUS, and it damn near burnt my finger. After installing a fan on the NB, it was an easy boot into WinXP.
I installed ASUS probe, and found that this mobo has a wicked VCore droop. 4.2GHz was failing in one instance of SP-2004 because of the voltage dip with both cores loaded. I had to set the BIOS VCore to 1.42v to get 1.34v under 100% load. I went on pass dual SP-2004 for 19 Hours , was satisfied with the stability, and stopped it.
It was now time to tweak the RAM, and I was surprised to see how many dividers were available. I worked my way up to 900MHz DDR, testing with Memtest86+. I ran though 10 loops of tests 5 and 8 until I reached 900MHz, then let all tests run for a few hours with no errors. This was the fastest I have ever seen this RAM run, and was very pleased to see an Everest memory read score of 8729 MB/s :Q
I am very happy with this upgrade so far....
I will be posting benchmarks and video encoding results in this thread (And my AutoGK Thread) as time allows. If there is something you would like to see run, LMK, and I will try to include the results. I have some Super pi results for now.... 
1M: 27s
2M: 1m 3s
4M: 2m 18s
8M: 5m 6s
16M: 11m 21s
32M: 24m 37s
EDIT: Observation: The P5WD2-P has a passive heatsink over the MOSFETs that depends upon air-flow from the processor HSF to cool it. When watercooling, (No air-flow) this heatsink gets extremely hot under load. :evil: I thought that this may be contributing to the VCore droop. After positioning a fan over it, I monitored ASUS Probe, and watched the VCore increase from 1.34v to 1.38v. Still drooping, but better nonetheless.
I was able to knock-back the VCore a couple of notches to obtain the 1.34v this current OC requires.
CrystalMark: 54,976
PCMark04: 8905
3DMark2001: 32892
AquaMark3: GFX: 14,877 CPU: 13,602 Total Score: 96,178
Cinebench 2003
ScienceMark 2.0 Benchmark Suite
POV-Ray Benchmark
I picked-up a retail P-D 920 and paired it with a refurb ASUS P5WD2-Premium mobo to lessen the $$ outlay.
Product Code: BX80553920SL94S
MM#: 878915
FPO/Batch#: L541B403
Pack Date: 12/30/05
I am using watercooling, which consists of:
Swiftech APOGEE waterblock
Swiftech Bay Res /MCP-350 pump
Swiftech RadBox external radiator mount
Heater core radiator
120MM Vantec Stealth fan with shroud
Other specs in sig....
I put the rig together, and went straight for 300MHz FSB... (4.2GHz) It stalled at the WinXP boot screen. :frown: I checked the NB heatsink on the ASUS, and it damn near burnt my finger. After installing a fan on the NB, it was an easy boot into WinXP.
I installed ASUS probe, and found that this mobo has a wicked VCore droop. 4.2GHz was failing in one instance of SP-2004 because of the voltage dip with both cores loaded. I had to set the BIOS VCore to 1.42v to get 1.34v under 100% load. I went on pass dual SP-2004 for 19 Hours , was satisfied with the stability, and stopped it.
It was now time to tweak the RAM, and I was surprised to see how many dividers were available. I worked my way up to 900MHz DDR, testing with Memtest86+. I ran though 10 loops of tests 5 and 8 until I reached 900MHz, then let all tests run for a few hours with no errors. This was the fastest I have ever seen this RAM run, and was very pleased to see an Everest memory read score of 8729 MB/s :Q
I am very happy with this upgrade so far....
1M: 27s
2M: 1m 3s
4M: 2m 18s
8M: 5m 6s
16M: 11m 21s
32M: 24m 37s
EDIT: Observation: The P5WD2-P has a passive heatsink over the MOSFETs that depends upon air-flow from the processor HSF to cool it. When watercooling, (No air-flow) this heatsink gets extremely hot under load. :evil: I thought that this may be contributing to the VCore droop. After positioning a fan over it, I monitored ASUS Probe, and watched the VCore increase from 1.34v to 1.38v. Still drooping, but better nonetheless.
CrystalMark: 54,976
PCMark04: 8905
3DMark2001: 32892
AquaMark3: GFX: 14,877 CPU: 13,602 Total Score: 96,178
Cinebench 2003
ScienceMark 2.0 Benchmark Suite
POV-Ray Benchmark