[jlmorton]

That doesn't pass the sniff test. The available voltage from a lithium ion battery will decrease as it is discharged. When your phone turns off, that does not mean there is no more energy in the battery. It means there is no longer enough voltage to power the phone.

Over the course of many discharge cycles, the battery will lose capacity, and the point when the voltage is no longer sufficient to power the phone will come sooner.

But this is overly pedantic. People generally consider this point to be simply an "empty battery."

Android phones do not suffer these performance changes. Instead, the phones lose battery life over time, and within a year or so, you might be lucky to get 12 hours of life out of a full charge.

You can make an argument that we should optimize for duration or performance, but the difference is that casual Android users are aware that their battery is deteriorating, while casual iPhone users believe their phone is itself deteriorating, or else much slower than the newer models.

[jtbayly]

Wrong. The peak demand from the phone can exceed the available voltage even if you're at 30%, causing an immediate shutdown. That's what we saw a year ago. Now they no longer shut down unexpectedly, but the performance of the phone sucks, even with a battery that holds 87% of its original capacity.

[jlmorton]

I don't dispute that, but that does not explain the observations that a fully-charged, reduced-capacity battery performs poorly on a standard benchmark test versus a new battery.

It's not just smoothing out a peak demand. It's reducing the peak performance all of the time, and that is unique to iPhone.

[esmi]

You're assuming the issue is only changing capacity vs age. Internal impedance increases with age as well. A full battery that can't source its energy in a timely manner might as well be empty.

[unobtaniumstool]

That's not entirely accurate. They observed that with a fully charged battery it performed as expected. The CPU wasn't throttled until the battery was depleted.

[jtbayly]

That's not my experience. Unless depleted means 95% charged and 87% capacity compared to original spec.

Mine runs at 50% CPU speed under those circumstances. I wouldn't call that "performing as expected."

[ric2b]

> The peak demand from the phone can exceed the available voltage even if you're at 30%

So throw in a fat capacitor. This was a solved problem 50 years ago, the only problem is this constant rush to make phones 1% thinner.

I've never met anyone that used phone thinness as a metric when choosing a new phone and yet here we are.

[valuearb]

Android phones do suffer from spontaneous reboots. It appears not as much as iPhones, and the most likely reason is the higher performance CPUs in iPhones. Android manufacturers have chosen to let the minority of users who suffer through spontaneous reboots, to keep suffering through them, rather than moderate CPU performance for them.