Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

andy06

macrumors member
Original poster
Feb 15, 2012
43
0
I seem to be getting especially slow boot times and short battery life on my new MBA13 (2011).

Boot time (From power on to desktop): ~18 to 20 seconds (Room mates older 2010 MBA13 boots in 12-14 seconds to desktop on SL, so different OS and different hardware, but both older)

Battery Life is about 5 hours (50% brightness, keyboard backlight turned off, no apps running except browser and constant multi tab browsing), basically replicating Apple's battery rundown test that they use to come up with the battery figures.

Sadly that is nowhere close to the 7 hours stated on their website or the 8-10 hours we can get out of room mates 2010 MBA13 running SL using identical workloads!

What are you guys getting and what should I realistically expect?
Thanks
 
I seem to be getting especially slow boot times and short battery life on my new MBA13 (2011).

Boot time (From power on to desktop): ~18 to 20 seconds (Room mates older 2010 MBA13 boots in 12-14 seconds to desktop on SL, so different OS and different hardware, but both older)
Performance Tips For Mac OS X

Battery Life is about 5 hours (50% brightness, keyboard backlight turned off, no apps running except browser and constant multi tab browsing), basically replicating Apple's battery rundown test that they use to come up with the battery figures.

Sadly that is nowhere close to the 7 hours stated on their website or the 8-10 hours we can get out of room mates 2010 MBA13 running SL using identical workloads!
There are many factors that impact your battery life. See the BATTERY LIFE FROM A CHARGE section of the following link for details.

This should answer most, if not all, of your battery questions:
 
Thanks for the links :)

My concern is slightly different though. The machine I'm talking about is new (no gunk, no battery cycles) and I'm getting consistent reproducible results of 20s boot times and max 5 hrs battery (when running our closest approximation of Apple's test). I have never exceeded 5.5 hours

However, similar units owned by others are getting the prescribed 7ish (6-8) hours of battery under identical load and my room mates OLDER MBA13 (2010) with SL is getting better life too with a 1 year old battery. It boots quicker too (couple of seconds).

I'm wondering if I picked up a Lemon :(
The boot time is largely academic but the battery life is crucial to me since some people seem to be getting 60% more battery life.
 
Thanks for the links :)

My concern is slightly different though. The machine I'm talking about is new (no gunk, no battery cycles) and I'm getting consistent reproducible results of 20s boot times and max 5 hrs battery (when running our closest approximation of Apple's test). I have never exceeded 5.5 hours
You can't know that if you have no battery cycles.
However, similar units owned by others are getting the prescribed 7ish (6-8) hours of battery under identical load
Except in tightly-controlled "laboratory" conditions, there's no such thing as "identical load" between computers. This is especially true between your new Mac and a year-old Mac. Read the links I provided that describe how to check processes that are running.
It boots quicker too (couple of seconds).
Again, look at the links I posted. They will address differences in startup times. Rather than skim over the links, take the time to read them and follow the instructions provided. You'll see your situation is no different than the thousands of others that have asked the same questions.
 
We started off with no battery cycles (which I checked using iStat, which may not be accurate) but obviously after testing it for about a month, its got a fair few cycles on it now (30)

I understand your point about the laxness in our pseudo-scientific tests and I read your link, but we *are* running pretty similar loads. Fresh boot, fresh installs, 50% brightness, KB Backlight Off, Same Wireless router, Same distance and then browsing (No Flash, No Java) on each till the battery drops to about 5%. Under these conditions, I have *never* exceeded 5.5 hours (mostly 5) and the other notebooks routinely go 6.5 to 8.5. So they're getting 50% more battery, consistently. That might be indicative of a problem, no?
We din't test it once or twice, more like 2 dozen times now.

Like I said, not worried that much about start up time. Was just surprised to see an older iteration of hardware + older iteration of software boot faster.

Thanks
 
We started off with no battery cycles (which I checked using iStat, which may not be accurate)
iStat is accurate.

Under these conditions, I have *never* exceeded 5.5 hours (mostly 5) and the other notebooks routinely go 6.5 to 8.5. So they're getting 50% more battery, consistently. That might be indicative of a problem, no?
So you mean that both computers are identical models with identical drives, RAM, OS versions, and have the exact same apps/widgets/processes/plugins/extensions running, right? And both have identical apps/widgets/processes set to launch on startup, right?
 
Yep, we tested 3 MBA 13" (2011) 128 GB SSD, 4 GB RAM, 1.7 Ghz Models, all running OS X Lion 10.7.2 till about 2 weeks ago and then 10.7.3 from that point on. There was also a 2010 MBA running 10.6 (so a total of 4 machines).

There are no widgets running when we do the test. There are no extensions and plugins (Flash, Java etc) installed on any of them. There are no start up items apart from iTunes Helper which was there by default I think.

We do it from a clean boot, so I think its a safe assumption that there wouldn't be too many disparate apps/processes (I din't launch any extra apps manually apart from Activity Monitor and Safari which run throughout the time).
 
We do it from a clean boot, so I think its a safe assumption that there wouldn't be too many disparate apps/processes (I din't launch any extra apps manually).
The only way to be certain about running apps/processes is to follow #5 in the performance tips.
 
I keep it on All Processes (Hierarchically) anyway. I like to always know whats going on with my machine. There was nothing that stood out. Obviously its impossible for both to be running exactly the same processes at exactly the same loads but since all of the machines are 'clean' (we bought em together) and we don't really clutter it with apps (the only 3rd party apps on all are VLC and Firefox, neither of which are used during the test). They really ought to be running more or less the same load profiles. Especially since the run times are consistent on a per machine basis.
 
I keep it on All Processes (Hierarchically) anyway. I like to always know whats going on with my machine. There was nothing that stood out. Obviously its impossible for both to be running exactly the same processes at exactly the same loads but since all of the machines are 'clean' (we bought em together) and we don't really clutter it with apps (the only 3rd party apps on all are VLC and Firefox, neither of which are used during the test). They really ought to be running more or less the same load profiles. Especially since the run times are consistent on a per machine basis.
If all computers are new, with the same amount of uptime, running the same processes in the same configuration, the times should be similar, but not exact. If they're significantly different, my bet is that you've missed something that's causing the difference, such as a process, Spotlight indexing, etc. that accounts for the difference in times. If you feel you haven't missed anything, take it to Apple.
 
I'll wait till a few people post their battery run times. If a lot of people seem to be topping around 5-5.5 no matter how hard they try, then I guess its just manufacturing variance. We don't have Apple stores in our country, so returning it isn't quite as easy, which is why I'm circumspect (sending it to them to take a look will cost me 2 weeks without the machine, at least) :(
 
I get about 5 hours with real world usage on my 13 inch MBA. The way to get more is to turn brightness down (biggest factor), and only using non CPU intensive software like safari or pages. Also turn off things like Bluetooth when not in use. also background stuff could help (I have Dropbox installed for example).

To really get to 7 hours you'd probably have to turn brightness almost all the way down.
 
Thanks Greg! I feel better knowing 5 hours is not awful! :p
The other machines seem to regularly do 6.5 on 50% brightness though. Looks like some batteries are better (bit like the Samsung/Toshiba SSD speed differences I guess).
 
Comparing battery life doesn't have to be so difficult. You don't need to sit around draining batteries with your friends ;)

Get Coconut battery. See what the current, design, and full values are. Then look at the current power draw. If the power draw is the same on all laptops, and the full capacity is the same, they'll end up giving the same battery life. If the current power draw is more on one of them, then try the same test on fresh lion installs. If the difference is still significant, say 7wh vs 9wh (which would result in 5 and 7 hr battery life), take it to an Apple store and have them diagnose it.

Your test of a few friends sitting around and doing internet browsing will always yield different results. Clone the exact same software on all, then watch the power draw.

FWIW, I get around 5 hours of real world usage vs 7 hours of lighter web browsing.
 
Thanks, will do that :D

What qualifies as Light browsing btw? Would 10 flash-less tabs in Safari still count as light? (Lets say I open up Appleinsider, middle click 4-5 articles and then read and close them as I'm done reading each one, is that light?). No flash, no animations, just reading. I never ever got more than 5.5 on that either. There was nothing else happening. That's why I was a bit surprised. I expected at least a very simple minimal case to nudge 7 hrs.

We don't have stores here, we have to mail it to them for a 2 week round trip. Hence the deliberation before shipping it :)

----------

I should probably mention by now that my first indication that something was wrong were not benchmarks from friends but online reviews (Anandtech, Verge, Engadget etc etc) where they usually run tests more demanding than Apple's test and still usually come to about the same ballpark figures. Apple has been very accurate with its estimates of late.

When reviews say they were listening to background music on iTunes, doing some light browsing, photo editing, had dropbox running and watched a few Youtube videos and STILL got significantly more than what I ever got, then time to get worried hehe
 
Yes that usage qualifies for light browsing, and I usually get 7 hrs of that usage (with brightness around 30 though, I also have spotlight only index my applications and settings, disable automatic keyboard and screen brightness).

With that being said, look at the math. The battery capcity is 50Wh. If your current usage in coconut battery (CB) shows up as being around 7Wh, then you'll get 7 hours (50/7). If it shows up at 10 Wh, you'll get 5.

Off a fresh install, no extra software, what is your power consumption in CB? With safari + flash/ad block?

Btw, any battery life estimates (say from Apple or a review) are usually based on automated tests. A google search will tell you that real world estimates are 3.5 and 5 hours for the 11" and 13" respectively. Generally its safe to subtract 1-2 hours from what reviews and companies tell you, because in real life you're doing a lot more than just browsing to sites. For e.g. you're using the trackpad and the keyboard, which is generating interrupts and not letting the processor sleep. The ambient sensor is manging keyboard and screen brightness and so on. Then if you're scrolling, smooth scrolling will use more CPU, so will gesture recognition etc. I don't know enough about the automated tests to know if they take all of this into account.

With that being said, if your idle usage on a fresh lion install is high (say 8 Wh or more), then you have an actual problem and you should send it in.

Edit: there's also a great tool called minibatterylogger (http://www.macupdate.com/app/mac/22649/minibatterylogger). It'll map your battery percentage over time. With normal usage I consistently see 20% drop in 1 hour. With light usage its around 15% per hour. Try it out.
 
My 2010 is a couple seconds faster than my 2011 at boot-up. Who cares? It gets smoked when I'm actually _using_ the computer.
 
Anandtech has found that the C2D (2010) 13" Macbook Air had an abnormally long battery life compared to the 11" of the same generation, as well as compared to the i5/i7 models from 2011. It seems unlikely that we'll see that type of battery life with either model using the new higher powered chips.
 
Andy, I know what you mean about similar workloads producing clear differences in battery drain between 2010 and 2011 MBAs, i've seen it repeatedly with friends' machines. The thing that worried me is exactly that you don't need a labs-tested, every-variable-controlled setting to notice it. It's one of the reasons I didn't upgrade to 2011.
 
Thanks guys! I thought I had some buggered unit (which is a huuuge deal for me coz we have no physical Apple support locations).

Anandtech has found that the C2D (2010) 13" Macbook Air had an abnormally long battery life compared to the 11" of the same generation, as well as compared to the i5/i7 models from 2011. It seems unlikely that we'll see that type of battery life with either model using the new higher powered chips.

Yeah, the 2010's profile was awesome. Though it may have been SL as it got worse after upgrading. The difference between the 10W Core 2s and the 17W i5s no doubt contributed.

The major difference however *may have* been from 50% on the brightness setting meaning different things on the 2010 and 2011 (Say 300 nits vs 400 nits). Screen Backlight trumps processor for power usage usually.

Have a look at this though (From macstories.net review)

I’m not the most indicated person to test a MacBook’s battery following Apple’s rules. During my average day, I keep brightness at 100% and I have 12 apps running in my dock (iTunes, Spotify, Twitter, Chrome, OmniFocus, Reeder, Mail, Evernote, Keyboard Maestro, OmniOutliner, Byword) plus a bunch of apps I may open and close from time to time. Additionally, I have other utilities running in the menubar or in the background such as:

Growl
ProwlMenu
Take Five
Dropbox
1Password
Cloud
Clipmenu
Coversutra
Alfred
LogMeIn
iStat Menus
Day One
TextExpander
Little Snitch
Hazel
Send2Mac

All of these apps have multiple processes/agents running in the background, and I tend to keep between 3-7 tabs always open in Google Chrome (which has Flash pre-installed and sometimes keeps even more than 30 tabs open for me). In spite of the MacBook Air performing just fine with these apps open all the time, battery life is clearly affected by the brightness settings I keep, and the software I leave open all day. For comparison, I thought it’d be interesting to see how much brightness alone could affect my MacBook Air’s battery life. So I ran two tests, one with brightness set at 100%, the other at 50%. The apps were the same (all those names mentioned above), and on both sessions I watched a 10-minute Flash video. Here are the results:

100% brightness: 3.5 hours
50% brightness: 5 hours

THAT.IS.INSANE. With his machine, I could get 10 hours out of that thing.

TL;DR: Roughly 30 apps open. Chrome with Flash installed and multiple tabs. He also watches flash video on it and still gets 5 hours on 50% brightness.

Andy, I know what you mean about similar workloads producing clear differences in battery drain between 2010 and 2011 MBAs, i've seen it repeatedly with friends' machines. The thing that worried me is exactly that you don't need a labs-tested, every-variable-controlled setting to notice it. It's one of the reasons I didn't upgrade to 2011.

THIS! I don't get the point of fretting over one process perhaps starting n stopping for 5 mins or other small differences (a widget here, a tab there) when the difference is so stark and apparent.

at the risk of being howled down, you could try running this one line at a time in Terminal

sudo chown root:admin /
sudo kextcache -system-prelinked-kernel
sudo kextcache -system-caches

What do they do :D
 
Ok will try em :)
Why would you get howled down for that?

And what's the counterpart to SMC on non-Macs? Like what are they called.

Thanks
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.