Measuring Devices and Testing Methods
by
Bert Toepelt
Testing methods for the processor and the complete system.
The processor’s energy intake is recorded directly at the electrical output of the voltage regulator. These measurements are not affected by the power consumption of the motherboard or any other components. To scale the processor’s performance loss, we measured voltage and current separately.
The complete system’s performance capacity is measured with the PM3000A high precision performance-measuring device from Voltech.
3
Comments
Sponsored
Latest CPU News
- 10/02 – CPU Performance Boosted 20% When CPU, GPU Collaborate
- 10/02 – Leaked Slide Shows Intel Haswell Set for March-June 2013
- 09/02 – $100,000 If You Can Prove Quantum Computers Impossible
- 09/02 – AMD Adds FM1 CPUs Athlon II X4 638 and Athlon II X4 641
- 08/02 – Raspberry Pi Scheduled to Launch This Month
my media centre based on a 3700+ (S939, single core san diego, 1MB, 2.2GHz) only uses 56-58W while running rosetta@home! That's including 1.25GB DDR (3 sticks), a freeview TV tuner, and a 2GB compactflash card on a Seasonic S12 330W PSU. It's undervolted as far as it'd go though...
What about performance per watt comparisons, which would have been the best comparison you could have made in such an article.
If you had a system that consumed 300W of power (forget about idle and full load differences for this question!!) with a 500W PSU, what would your power consumption be for the purpose of energy bill calculation?
Am I correct in believing that the rating of your PSU is the maximum power it can supply, and that it only actually draws what the system asks for? So in this case, your overall system power use would be 300W?
So, if you install a much more powerful PSU than you currently need (for the sake of future SLI upgrades) you wouldn't be wasting electricity?