It would have been interesting to see under-volting data using RMClock. Intelâ€™s 45nm CPUs donâ€™t under-volt as significantly as the 65nm ones so the lack of BIOS under-volting is less of an issue.
BIOS under-volting is more useful if youâ€™re under-clocking as the 45nm dual-cores seem to start at ~2.5GHz and at that speed the voltage doesnâ€™t seem to be stable much lower than is available with software.
Itâ€™s quick to test RMClock with the entry level 45nm dual-cores as they are typically stable at the lowest available VCore so you only have to test stability using one setting.
While these results are profound, we won't be changing our Intel motherboard test bed to include the E7200 as getting the absolute lowest system power consumption isn't our main testing goal. A slower processor is ideal for judging an IGP's video playback efficiency â€” if the E7200 ran at around 2GHz, we would not hesitate to change our methods.
2.5GHz is becoming the new entry point for Intel dual-cores although there are still older chips around that are slower and cheaper.
The E5200 is worth considering as even though it runs at 2.5GHz you can easily restrict it to 2GHz by limiting the maximum multiplier to 10 using RMClock if you want to test at a lower clock speed. Although whether you test video playback with the clock speed at 2 or 2.5GHz seems largely irrelevant as the CPU load percentage in conjunction with the clock speed should tell you how much processor power is being used.
So while we won't be changing our test bed, this little experiment gives us something to keep in mind: AMD is not the undisputed king of CPU power efficiency any longer.
And hasnâ€™t been for almost two and a half years now.