devx, that board doesn't have gigabit LAN. The S2668AN
R model does for $298. You're also still missing 512MB of RAM. So the difference is still over $300. There is no way you can make a Xeon system cheaper than an Athlon MP system because everything else created equal, the core components of CPU's and motherboard both cost more on the Intel side. I won't argue that the quality is better on the Intel side because it is, but again, we're supposedly building systems that you can put under your desk and use at home, not corporate servers (why would you put an Audigy 2 Platinum and Radeon 9800 Pro in a server?). For business I would use nothing but Intel. At home, the extra cost doesn't get me anything worth having. The goal was also to get the cost under $2999 which was the cost of the base model dual 2Ghz G5.
BoberFett, that's a great deal on that Tekram card. They're also selling a Tekram U320 card which I didn't know existed, at a cheaper price than the U160 I had listed, so I could still squeeze a U320 controller in and keep the total tag under $3000. Interesting that it has a fan on it. Never thought I would see that.
Final Specs:
Lian-Li PC-7 -> $96
Antec TruPower 480W -> $87
AMD Athlon MP 2800+ x 2 -> $494
MSI K7D Master -> $188
1.5GB Crucial Buffered/Registered (3x512MB) -> $354
ATi 9800 Pro -> $389
Audigy 2 Platinum -> $167
Tekram DC-390U3W -> $80
Fujitsu MAS3184NP (18GB 15k) -> $187
Seagate 7200.7 ST3160021A x 2 (320GB total) -> $262
Plextor PX-708A (8x DVD+R, 4x DVD-R) -> $260
Intel PRO/1000 MT Server Adapter (PCI-X gigabit LAN) -> $102
Firewire 800 card -> $70
US Robotics V.92 Modem -> $36
Logitech Cordless MX Duo -> $78
XP Pro -> $143
Total -> $2993
Was that for a reason? Do they know something that we don't? or they just didn't have one readily available?
He posted somewhere else that he only benchmarked what he had available to him, which is why nothing faster than single P4 2.66GHz system was tested on the Intel side.
Here's the link from Arstechnica:
http://arstechnica.infopop.net/OpenTopic/page?a=tpc&s=50009562&f=48409524&m=3330955475
"The main goal was to evaluate the G5, and use some of our existing G4 (1GHz and 1.25GHz) and P4 (2GHz and 2.66GHz) systems as reference. I was testing on machines used in my work and comparing to a new system that was the focus of the report. The G4 and P4 machines in my study are 1-2 year old technology, but that's what we have to compare against."