The higher binned processor of the same generation (with the same TDP) is always the more energy efficient choice. Because they require less energy per click cycle.
e.g. 4.8Ghz at 28W, vs, 4.2GHz at 28W
You’re able to get more clock cycles per given energy expensed on the faster processor.
I’ve stated that in some of my other posts.
Depends on where you stand on the cost difference…to some the performance difference (and therefore energy efficiency) doesn’t warrant the price difference. Comes down to your personal view of ‘value’.
Curious, why do you want to upgrade if your use is light, and you can downgrade the CPU via the BIOS.
So the argument can be buy a really high end as they are more efficient then downgrade it via the bios, but it won’t make a lot of difference to the battery unless you measure form one extreme to the other, and as light user it probably won’t make a noticeable difference.
Of course how you charge and discharge may have more effect on battery life that the small changes on load.
Not even close, but you’d probably have to read more deeply into the thread. Try searching for thermald, auto-cpufreq, and rapl to see what you could be doing if you actually wanted to max out battery life. It’s pretty easy to set arbitrary power usage limits (which would of course give you extended battery life).
As I mentioned in the thread inquiring about 12th gen BIOS options, it’s theoretically possible, if one is willing to sacrifice out-right performance, that disabling as many P-cores as possible could substantially increase battery performance by virtue of forcing more processes to run on the E-cores rather than the P-cores.
Again, this is all just theoretical.
…speaking of which, in such a situation, I wonder if disabling hyperthreading would help or hurt battery life, e.g. would two process threads distributed as 1 P-core thread + 1 E-core thread consume more or less battery compared to 2 P-threads on a single P-core?