Originally Posted by unclewebb
The Asus board I am using ignores the EIST request in the bios but you can disable EIST using software like RealTemp after you boot up. If you are using C6, there is no need to disable EIST. It doesn't make any difference so you might as well leave it on.
If you are seeing an 8 watt difference then you need to look at your testing a little closer. Before making any comparisons, make sure your computer has had a chance to settle down after booting up. Depending on what you have running on your system, this might take a while. Keep a close eye on the RealTemp Load % to make sure it is low and not fluctuating so you can make a fair comparison between different settings. Some programs only start working when the CPU is idle. If you have programs like that running in the background, it might be difficult getting repeatable and comparable results.
Post a RealTemp and RealTemp -C States window when your CPU is idle like I posted.
If you are using the C Package states, disable them when doing your initial testing.
The low power C6 state is like pulling the plug and disconnecting your house from the power grid. If your cores are spending over 99% of the time in C6 at virtually zero volts, no change in voltage should result in an 8 watt difference in power consumption. What are you using to measure power consumption?
I did test them when the computer is stable and the results were repeatable. I closed all apps and services running to make an accurate reading.
I will post a screenshot later.
There is no C Package states setting in my BIOS so I assume they are enabled by default because the Package State for C6 in RealTemp shows that my CPU is in Package State C6 65% of the time when idle.
Exactly. I was expecting no change in result between Offset and Fixed vcore as long as C6 is enabled. This makes me wonder why Offset has lesser idle power consumption given they virtually have the same close-to-0 voltage. Maybe, the current draw is more in Fixed vcore than in Offset accounting for the 8-Watt difference. I am using an APC BR-1500GI UPS to measure power consumption and I'm pretty sure this is accurate like a Kill-a-Watt meter.
My big question though is why do you think Fixed vcore is better than Offset? When you do any task in a computer, the CPU rarely stays at 1.6GHz anyway but it is still better that at idle it goes to 1.6GHz. Or are you saying that the only reason that Fixed vcore is better than Offset is because you can make your system stable (when overclocking) easier with it?