Overclocking
We kicked off our overclocking tests by dropped the CPU multiplier to 6X—its lowest possible value. We also reduced the memory bus speed to keep our DIMMs operating within their limits. Next, we turned our attention to the front-side bus on each board, cranking it up and using a two-way Prime95 load to test stability along the way.

The P5Q3 Deluxe sailed all the way up to a 500MHz front-side bus without so much as batting an eyelash. We had to add the auxiliary chipset fan to get the board stable at a 510MHz FSB, and that was all she wrote. Applying a little extra chipset voltage managed to coax a 520MHz front-side bus into Windows, but our stress test quickly spat out errors or crashed the system. Additional voltage wasn't of any help, either.

Like the P5Q3, the P5Q Deluxe sailed up to a 500MHz front-side bus speed without issue. Our board would go no further, though, even with extra voltages and the auxiliary chipset fan.

Given that we're dealing with early samples, the fact that neither P5Q had problems handling 500MHz front-side bus speeds with stock voltages and passive cooling certainly bodes well for the P45 Express' overclocking potential. The P35 was no slouch in this department either, so I suppose we shouldn't be terribly surprised.

In addition to pushing the front-side bus speeds, we also took a stab at memory overclocking. The P5Qs sport a feature called Memory OC Charger, which according to Asus, is the culmination of efforts to improve DRAM signal quality to enable higher memory bus speeds. Asus has a list of OC Charger-validated memory modules that includes DIMMs from Corsair, OCZ, Kingston, GeIL, and others, and our DDR3 modules just happened to be on that list.

The DIMMs are only rated for operation at up to 1600MHz, but we managed to push them to an effective 1840MHz on the P5Q3 Deluxe. Impressive.

For the P5Q, we swapped in a pair of Corsair Dominator DIMMs and were able to take them up to 1200MHz—89MHz beyond their rated speed of 1111MHz. Not bad at all.

Power consumption
We measured system power consumption, sans monitor and speakers, at the wall outlet using a Watts Up Pro power meter. Power consumption was measured at idle and under a load consisting of a multi-threaded Cinebench 10 render running in parallel with the "rthdribl" high dynamic range lighting demo.

In the graphs below you'll find two instances of the Gigabyte X48T-DQ6 and the Asus P5Qs. We ran the Gigabyte board with and without its Dynamic Energy Saver (DES) feature enabled. The P5Qs were also run with and without their Energy Processing Units (EPU) enabled. We settled on Six Engine's high performance energy savings mode since it's the only one that doesn't reduce system performance.

Even with loads of onboard peripherals, the P5Qs still manage incredibly low power consumption at idle and under load. The idle results are the most impressive, putting the boards a good 10W clear of their closest competition.

Considerably less impressive, however, are the EPU results. Six Engine's high performance mode doesn't do much for system power consumption at idle or under load. Dropping to low power mode does reduce idle and load power consumption to 96W and 140W, respectively, but that's to be expected from a slower configuration. We also found that the EPU energy saving scheme wasn't entirely stable on our P5Q3 Deluxe. The P5Q Deluxe was fine, though.