But we're talking about something used to surf the web. An ARM is still powerful enough and will use way less power.

Web browsing is pretty darn good on the iPad and comparable Android tablets, but it's still a far cry from the performance you can get on x86 hardware. Many Surface Pro reviews mention this specifically. Now, there's still obviously value judgement to be made: is the Pixel worth the extra money, extra weight, and worse battery life compared to an Apple or Android tablet?

Worse battery life compared to an Apple? I think you are wrong. I think Google is being pessimist when estimating the battery life to be 5h when their capacity is 59Wh. For comparison, Apple advertises the MacBook Air 13" battery to be 40% longer (7h) when their capacity is 15% less (50Wh). Maybe Google estimated a higher average power consumption (eg. using the LTE wireless network which is more power consuming than wifi).

  MacBook Air 13": 50Wh - starts at $1200
  Pixel: 59Wh - starts at $1300
  MacBook Pro 13": 74Wh - starts at $1500

Basically, the more you pay, the more capabilities and battery life you get. And both Apple and Google seem in line.

> I think Google is being extremely pessimist when estimating the battery life to be 5h when their capacity is 59Wh.

The battery life depends, not on the watt-hour rating, but on the watt-hour rating divided by the power required. More power, fewer hours. Maybe Google knows something about the actual power requirements that resulted in this estimate.

Just as an example, if the unit requires 11.8 watts, then the battery life will be:

59 watt-hours / 11.8 watts = 5 hours

> For comparison, Apple advertises the MacBook Air 13" battery to be 40% longer (7h) when their capacity is 15% less (50Wh).

Different battery, but also different power requirement.

> Maybe Google calculated it while assuming you were using the LTE wireless network, or streaming HD movies, or something else quite power-hungry.

You're speaking as though all these units have the same power requirements.

I am well aware that the battery life depends on the power required. My point (implied) was that the MacBook Air 13" has roughly the same hardware as the Pixel: same CPU, same GPU, same amount of RAM, same type (DDR3L), etc. The only significantly more power-consuming devices in the Pixel are the display and LTE hardware. So if you don't use too much these (eg. turn brightness down, don't use LTE), then shouldn't the Pixel be expected to consume about the same amount of power as the MacBook?

Let me reply this way -- why would Google sell their unit short? What reason would they have to underestimate battery life?

It occurs to me that, if they exaggerated battery life in normal usage -- typical user brightness settings and typical peripheral activation -- it might create a public relations disaster not unlike the infamous Tesla test drive.

Also, even with the same battery technology, units won't necessarily behave the same as the battery voltage falls. Some will cut out sooner than others as they approach the end of battery charge, a trait that depends on the power supply design.

On general principles, I just don't think they would intentionally undercut their own unit.

Everybody knows there is no industry standard to measure battery life (basically which workload to run when measuring it). For example Anandtech measured the MacBook Air 13" as having a battery life varying between 3 and 7.52 hours depending on the workload: http://www.anandtech.com/print/6063

So my point is that Google would not specifically have the goal of underestimating battery life, but just that it is almost guaranteed that they have different standards from Apple for measuring battery life.

The only way to fairly compare battery life is to have the same person/group run the same test on the 2 laptops. But even that is prone to human-error (eg. what if the tester forgets to disable LTE when testing the Pixel?).

> ... it is almost guaranteed that they have different standards from Apple for measuring battery life.

Yes, most likely, which is why Consumer Reports articles are often such interesting reading.

> The only way to fairly compare battery life is to have the same person/group run the same test on the 2 laptops.

Yes, and (just to make it more scientific) the testers shouldn't be able to tell which laptop they're testing. Not so easy to do, but necessary for an unbiased evaluation.

I just think Google's battery life specification is most likely correct -- they have no good reason to over- or underestimate it. But I agree that an independent test would be useful.

The battery life testing mechanism for Chrome OS is documented here: http://www.chromium.org/chromium-os/testing/power-testing.

Thank you, this confirms what I suspected. Google makes the laptop stream music and play full screen video for some amount of time, whereas Apple makes it only browse wirelessly ("up to X hours wireless web").

The screen and the greater graphics usage related to it is likely to be significant. It may draw much more power for any given brightness than an MBA.