Hacker News Reader

Comment by mathieuh | Hacker News Reader

I'm not experiencing any real change in battery life on my specced out 15" retina. Got ~5 hours before 10.9, get ~5 hours on 10.9. Makes me kind of wish I'd gone for the base CPU and saved money and battery life.

View on HN

Is there a difference in battery life? The CPUs all rate the same maximum wattage (it's the same chip after all!) and CPUs are pretty good about not using power when idle.

Not to mention that a faster clocked processor has to stay active for less time to achieve the same task. A similar logic explains why LTE battery life is often better than 3G/UMTS too.

Unless I'm mistaken, isn't something like 75% of a CPU's power consumption just in pushing the clock signal around? I just assumed that a faster clock would consume more power (as a general rule).

Wikipedia seems to agree with me: http://en.wikipedia.org/wiki/CPU_power_dissipation

You CPU doesn't have to run at the full GHZ when it isn't in use.

Yes a faster clock will do that, but as mentioned by quellhorst, you won't always be running at the fastest speed. There's also another important trick Clock Gating[1] that will reduce the power even more. That's one of the places that Intel has been working on in order to get the power usage down as much as possible to make the CPU have to dissipate less heat too. With a multi-core system it's just about absolutely necessary in order to be able to avoid overheating the die.

[1] http://en.wikipedia.org/wiki/Clock_gating

Are you sure?

http://gigaom.com/2012/02/17/why-lte-sucks-your-battery-that...

May be worth running this: http://software.intel.com/en-us/articles/intel-power-gadget-... and seeing if any software you use is pegging it to maximum unnecessarily. I find that Microsoft Silverlight on MacOS tends to do that; Flash also does occasionally, though not always.

Reply on news.ycombinator.com