Why higher power usage? i7-2630qm vs i7 3630m

[Enigmoid]

I had an old dell xps 15 laptop which used the i7 2630qm processor which I replaced with an newer lenovo y580 (i7 3630m) (because it overheated too much). The dell xps at idle usually used about 3-4 watts of power on the cpu, the lenovo uses about 4-6 watts. Both are at idle on maxmimum battery modes. Power measured using hw monitor. I thought ivy bridge used less power than sandy.

 

[ShintaiDK]

You cant compare different laptops like that.

 

[Enigmoid]

You cant compare different laptops like that.

why not?

Aren't I looking at cpu only?

 

[dma0991]

why not?

Aren't I looking at cpu only?

Too many variables to take into account. Both of the CPUs you've mentioned have the same TDP but with the IB clocked higher, nullifying the power saving for more performance. Then there is with different motherboards. They work differently in terms of its efficiency, you'd have to take that into account as well.

Lastly, HWMonitor will not give you the most accurate reading. I'd take 2W difference into its margin of error given that it is a software based reading. Whats more important is that you take into consideration the entire power draw of the whole system rather than focusing on the CPU alone.

 

[Haserath]

why not?

Aren't I looking at cpu only?

Different laptops are not comparable, especially if you're only comparing idle power. The laptops have completely different components besides just the CPU.

Both Ivy and Sandy will idle around the same wattage. Loads are different, but that also will depend on clocks, cores, etc.

Plus, HWmonitor sucks for power readings.

 

[Enigmoid]

Thanks, i didn't know that hw monitor was so inaccurate.

Is there any way to reduce idle clocks on ivy bridge?

 

[RU482]

did you mean i7-3630QM???? I don't see an i7-3630M

 

[Enigmoid]

did you mean i7-3630QM???? I don't see an i7-3630M

yeah, sorry

 

[tweakboy]

Isn't Haswell gonna have names like i7-3760k and so on... hmmmm

 

[Eureka]

Don't go by idle readings. Idle power is more affected by peripheral usage than load power.

To really figure out power draw, I would disconnect the battery and run off of just the power brick. Use a kill-a-watt at the wall as well as HWMonitor. Run Intel Burn Test, Prime 95 or whatever else you have to tax out your computer. Eliminate as many sources of error as possible. I would even say, don't measure the total power but rather the power difference from idle to load (elminates variations in screen, motherboard, peripheral usage, and power losses in the system).

 

[kevinsbane]

The largest power draw at draw in current laptops is the backlight and driving circuitry for the laptop monitor.