Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Doesn't a LiIon battery require a few charge cycles before it gets up to optimum capacity? I'm not really surprised that the numbers are so low out of the box.

After 4 or 5 charge cycles, it would be interesting to run the benchmark again.

Oddly the reviewers themselves often mention this, but we forget when we read it.
 
lol, 20 pages in a loop, so browser only read from cache? wifi not in use after all?
It depends on the pages they are visiting. If they are news site pages they would change fairly often and require downloading... or those with ads that are rotated, etc.
 
Incomplete and almost useless testing

How come they didn't test the use of bluetooth and usb and the external usb cd/dvd drive and isight camera. The air already has a few i/o's and they couldn't test those. It would be nice to know how much power these things would suck even for a moment of use. People do sync their iPhone and other iToys with iTunes.
 
Most people complaining about how unreal the 'real world' tests are are making the simple mistake of thinking that the reviewers are even vaguely interested in the results and how pertinent they are to real users. They are journos! - what they are interested in is scamming and blagging the latest product - the MBA in this case - and showing it off to their mates. They are certainly not going to waste valuable bragging time by giving you a comprehensive real world test - what do they know about the real world anyway?
 
Most people complaining about how unreal the 'real world' tests are are making the simple mistake of thinking that the reviewers are even vaguely interested in the results and how pertinent they are to real users. They are journos! - what they are interested in is scamming and blagging the latest product - the MBA in this case - and showing it off to their mates. They are certainly not going to waste valuable bragging time by giving you a comprehensive real world test - what do they know about the real world anyway?

Yes, how dare they make any negative comment about an Apple product! Tar and feathers! :rolleyes:
 
Plan ahead

The lack of control between different tests and different brands has been pointed out.

However, skimming the comments, I see little attention being paid to simple planning ahead. Regardless of the computer, If I'm on battery and away from a power source, I take that into account at all times. Clearly, you kill wi-fi, and bluetooth plus lower the screen brightness to an absolute minimum every time you can. Taking a few minutes to talk on the phone, kill the screen or even sleep. The point is that I'll end up with 2x the battery life of my wife who can't be bothered.

The long and the short is that I have NEVER run out of juice since I bought my iBook or with my current MacBook. I suggest 90% of people could accomplish the same if they stayed vigilant.
 
Most people complaining about how unreal the 'real world' tests are are making the simple mistake of thinking that the reviewers are even vaguely interested in the results and how pertinent they are to real users. They are journos! - what they are interested in is scamming and blagging the latest product - the MBA in this case - and showing it off to their mates. They are certainly not going to waste valuable bragging time by giving you a comprehensive real world test - what do they know about the real world anyway?

They wouldn't be very good reviewers if they 'scammed and 'blag'. Quite a broad and overreaction IMO.
 
The battery is pretty good. I would expect more but hey...its a Rev. A. I am sure the battery will be great in later models.
 

I did a quick search and it looks like I jumped the gun a little. There are external battery packs using Magsafe for MB and MBP and at least one supplier has said that they'll be releasing an external battery pack for MBA soon.

So it's not available yet, but it will be - for the infinitesimal number of people who need more than 5 hours away from any power outlet, car power adapter or airplane power adapter.
 
Honestly, I don't see what people are crying about. These are perfectly acceptable tests to determine battery life under light, heavy medium and heavy loading.

It's not a personal attack, you know. It's just a fact of life - the amount of life you get out of your notebook battery depends on what you use it for and Steve Jobs' figure of 5 hours is an optimal figure - i.e. the best it can get. Every other laptop manufacturer does the same thing and all their battery tests show the same thing - less battery life under heavier loading.

People should chill out a bit. I don't think anyone really believed the Air had a 5 hour battery life in reality and these figures aren't actually that bad.
 
what usage do you get 5 hours from MBA?

Some people are reporting 5 hours. Ask them.

But Anandtech gets 4.27 hours with continuous web usage, continuous hard disk access, continuous mp3 decoding, continuous wifi and high screen brightness. It isn't too hard to believe that more reasonable usage will get you 5 hours.
 
Fine

Honestly, it's no big deal. If you're a "heavy" user, and on a trans-continental flight, just bring along a second battery to pop in.

Oh wait. Never mind.

Guess I'll be enduring those whole two extra pounds of the MacBook in my carry-on bag after all.

Gosh, I hope I can find room for that extra quarter-inch of THICKNESS! :rolleyes:
 
Gas mileage is the same thing. Give it a try. Your car may advertise 30 hwy, 25 city. But guess what? That's in a vacuum going downhill. :)

In real life you get maybe 75% of that. More like 24/19. Supposedly the numbers are going to be real world this year with new specs by govt on gas mileage.

One problem is that the "hwy" (I suppose it means "Highway") test cycle actually runs at speeds between I think 25 and 45 mph. I'd love to have a number for running at continuous 70 mph as well.

And that brings me to the subject of benchmarks: Cars will be optimized for this benchmark, not for real-life usage. To optimize for the official test, I would adjust the gears so that you can just about drive smoothly at 45 mph in the highest gear. This will give you best fuel consumption in the benchmark. It will increase fuel consumption at 70 mph because at that speed your revs will be higher than necessary.

Same can happen with computer benchmarks. Apple has often been criticized for example for "bad thread performance" in some benchmark, and keeps replying that they optimize their code for real-life tasks, not for benchmarks.
 
I got more than 5 hours the other night. I was working exclusively in iwork(pages), switched wifi on and off a couple of times to upload files/check email. Screen was dimmed yet bright enough for me since I was in a darker room.

Is this a "real world" scenario? Probably not-but neither are the tests where they're listening to music while editing a movie while internet surfing with wifi while downloading 100mb files until the battery empties out.
 
And that brings me to the subject of benchmarks: Cars will be optimized for this benchmark, not for real-life usage. To optimize for the official test, I would adjust the gears so that you can just about drive smoothly at 45 mph in the highest gear. This will give you best fuel consumption in the benchmark. It will increase fuel consumption at 70 mph because at that speed your revs will be higher than necessary.

Whilst I completely agree that all manufacturers optimise as much as they can get away with for tests, I believe for mileage tests on cars the manufacturers must use the cars as they are when they come off the line and go for sale. That is, they aren't allowed to adjust the gears any differently to the ones they sell.
 
Some people are reporting 5 hours. Ask them.

But Anandtech gets 4.27 hours with continuous web usage, continuous hard disk access, continuous mp3 decoding, continuous wifi and high screen brightness. It isn't too hard to believe that more reasonable usage will get you 5 hours.

I got 4:13 in testing on the very first full discharge (cal cycle). Web page refreshing every few minutes, continuous audio play in iTunes, Mail checking every few minutes and about an hour of intermixed Screen Share. Display brightness at about 60%.

I felt this was very close to, if not better than Anandtech's (considering the screen sharing overhead, both CPU & wifi).

This was actually better than I expected, especially for the very first cycle.

I'll try another one in a couple of days (have to give my wife some time on her new machine, after all ;) ).
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.