It would seem that Xbench is overly weighted toward ONE processor in its ratings and doesn't take proper advantage of multiprocessor machines, making this app somehwat worthless when comparing machines. Why else would a *DUAL* 2 GHz G5 be only a few percent faster than a *SINGLE* 1.8 GHz G5? For anyone who doubts this, install CHUD, turn off one processor on your dualie and run the benchmarks...CPU test: 193.10 (dual 2GHz G5), 182.06 (dual 2GHz G5 w/one processor disabled with CHUD). The real world doesn't suffer this anomaly.