ScienceMark 2.0	Source: ScienceMark

Mol Dyn: Molecular Dynamics is a method for simulating the thermodynamic behaviour of materials using their forces, velocities, and positions. Most important of the afore mentioned is the force.

Primordia: This code calculates the Quantum Mechanical Hartree-Fock Orbitals for each electron in any element of the periodic table.

ScienceMark 2.0 Benchmark Results
Mol Dyn	Seconds	Ranking
Athlon64 3200+	81.34
Pentium 4 3.2C	80.49
Dual Xeon 3.06 GHz HT Enabled	72.48
Dual Xeon 3.06 GHz HT Disabled	90.57
Single Xeon 3.06 GHz HT Enabled	109.40
Single Xeon 3.06 GHz HT Disabled	143.23
Primordia	Seconds	Ranking
Athlon64 3200+	420.24
Pentium 4 3.2C	402.75
Dual Xeon 3.06 GHz HT Enabled	414.05
Dual Xeon 3.06 GHz HT Disabled	423.48
Single Xeon 3.06 GHz HT Enabled	450.66
Single Xeon 3.06 GHz HT Disabled	560.55

The dual Xeon system finally comes into its own here. Sciencemark 2.0 makes good use of hyperthreading and multiple processors, and this shows especially in the Mol Dyn benchmark. It's nice to see that the program can differentiate between dual CPU's and HyperThreading as well. Bandwidth is king in the Primordia benchmark, and the 800MHz FSB Pentium 4 takes advantage of this.

CINEBENCH 2000	Source: Cinebench

CINEBENCH 2000 is a free benchmarking tool for Windows and the Mac OS based on the powerful 3D software CINEMA 4D RX. The tool is set to deliver accurate benchmarks by testing not only a computer's raw processing speed but also all other areas that affect system performance such as OpenGL, multithreading, and multiprocessors

CINEBENCH 2000 Benchmark Results
Multiple CPU	Rating
Dual Xeon 3.06 GHz HT Enabled	62.52
Dual Xeon 3.06 GHz HT Disabled	51.29
Single Xeon 3.06 GHz HT Enabled	34.82
Single Xeon 3.06 GHz HT Disabled	N/A

Cinebench is the first of our benchmarks (besides Sandra) that specifically tests multiple CPUs. We thought we'd run the Xeon system through the multiple CPU benchmark to see how it did in its various configurations. We knew that hyperthreading allows single Intel CPUs to run this multiple CPU test, so we wanted to see the difference between Hyperthreading's 'virtual' extra processor and a physical dual CPU setup. As you can see, there's quite a significant performance advantage to the extra CPU here, though hyperthreading alone allows a single CPU to do a competent job.