Power Consumption
Here’s where this discussion really comes to a head. If you could tweak any of these chips up to 4 GHz without affecting any other variable, we’d already have our recommendation based on the results of our benchmark suite. However, that’s not the case.
The good news is that you can increase the voltage of either processor, crank them up to 4 GHz, and still end up with a fairly tame idle power result. With Enhanced SpeedStep properly implemented on Intel’s DP55KG motherboard, even keying in a 200 MHz or 190 MHz BCLK setting means that both of our test CPUs scale back when they’re not being used. There’s a minor power penalty in both cases, but at two or three watts, it’s easy for us to ignore.
A chart of a PCMark Vantage run on Intel’s Core i5-750 tells a drastically different story when there's a load applied, though. You’ll find three lines on the above chart: the green one represents our run with Turbo Boost disabled completely, the red one shows power with Turbo Boost enabled, and the blue line is power with the platform overclocked to 4 GHz using a 200 MHz BCLK and 1.45V core setting.
Clearly, turning on Turbo Boost exacts a power consumption penalty. But it’s much smaller than the result of cranking up the settings needed to make 4 GHz a stable setting on a 2.66 GHz processor.
The average across the entire run with Turbo Boost disabled is 115W. With Turbo Boost turned on, the average steps up to 120W. Overclocked to 4 GHz, the average jumps to 156W, and we’re only finishing the test 28 seconds faster.
