Idle/Peak Power Consumption Results
Idle Power
We don’t have a chart showing the idle power consumption of the AMD Phenom II X4 955 because idle power doesn’t change on AMD processors. Once Cool’n’Quiet is active, it will always set the processor to 800 MHz core clock speed and 0.96 V in the case of our MSI 790FX-GD70 motherboard. Hence, the system always required the same 99 W idle power.
The chart shows the system idle power on the Core 2 Extreme QX9650 system at all the voltage levels we tried. The 87 W system idle power could be reached at a voltage of less than 1.1 V, while the 1.255 V default resulted in 94 W idle power.
Peak Power
The power savings through undervolting are significant in the case of AMD’s flagship processor. We started at the default voltage of 1.32 V, which caused a 216 W system peak power, and we finished at only 1.175 V and 179 W peak power. The power savings of 17.2% or 37 W are sufficient to run an additional 20” display within the level of power consumption this system shows today!
Can the Intel system beat the 17.2% reduction in peak power? It can: the quest for the lowest reliable processor operating voltage here resulted in 1.078 V instead of 1.255 V, and a system-wide power consumption of 148 W instead of 185 W—that’s a 20% reduction.
Latest Tweaking & Tuning News
Latest Tweaking & Tuning reviews
- 11/05 – Do-It-Yourself: Upgrading Apple's 27" iMac With An SSD
- 17/04 – Overclocking: Asus Rampage IV Extreme Versus EVGA X79 FTW
- 30/03 – System Builder Marathon Bonus: Newegg Customer Choice PC
- 29/03 – System Builder Marathon, March 2012: System Value Compared
- 28/03 – System Builder Marathon, March 2012: $2600 Performance PC
Half-hearted article. If you 're serious about doing this (or you have a laptop) try RMClock.
Good article.
Half-hearted article.
Since the topic is power saving it is understandable.
One addition: for the readers not able to measure the power saving , you can check the CPU temperature with and without the voltage modification and you will see the difference.
A second addition: a true power saver should also accept a slight performance drop. So lowering peak frequency is another option.
AMD finally managed to threaten Intel’s Core 2 Quad in some benchmarks, but at somewhat lower cost both for the processor and an average platform. However, the Core i7’s reign remains untouched.
Actually, AMD beat the s**t out of C2Q and threatened I7 in some benchmarks. Get it right Tom. Darn I hate this site.
Where did it beat the shit out of C2Q, performance per watt intel win.
Does anyone think there would be such margins for undervolting notebooks or would they be tuned a bit better than their desktop counterparts to maximize battery life anyway? Of course in most cases this is a hypothetical question since I believe that most notebooks don't allow alteration of processor voltages up or down.
Amd4life quit being a fanboi. You're not any cooler for being a hater either. Intel is beating the crap out of AMD performance wise it's just the way it is. Get over it..
Amd is beating the crap out of intel pricing wise.. It's just the way it is....
Fanbois only hurt themselves.. You miss out on so many good products because you think you're being loyal when in reality you're only being a retard!
AMDFangirl has been undervolting stuff and posting about it for 18 months ... and quite a few have been undervolting the 45nm Intel quads in the forums too.
Why not have the best of both worlds
I have OCed my I7 to 3.3Ghz while reducing the voltage to 1.1V (after vDroop)
and my I7 is not that great, i have seen some that do the same OC with much less
I had an Opteron 165 Dual Core 1.8ghz stock v1.35 ocd to 2.7ghz at v1.2 ! Incredible...
AMD4Life, i would like to see some facts or benchmarkscores of that, the i7 remains untuched and the amd and intel duo is splitting haris performance wise
The intel is more efficent but the amd is more secure for the future.
The i5 will probably beat the heck outta the amd in performance at the same price and also being secure for the future.
my 2 cents, split them either way.