I realize you always have to take MB sensor readings with a grain of salt, but it appears that my SD's CPU temps actually dropped by ~ 4C at idle and under load simply by switching from the 180NFM(Rev.3)1T bios to the 136NFM(Rev.4)1T bios while - get this - running 0.05 more volts (1.65 vs. 1.60) than the 180NFM at a higher HTT frequency. Can this be even remotely possible? If it isn't just a misreporting bug, I might try bumping my VID a little further to see whether I can wring a few more MHz out of this chip.