MBA 2010 vs. 2011 vs. 2012 temps

[AtmChm]

Does anyone have any data on average temps for the MBA models of the last three iterations? I have a 13" and 11" 2010 MBA, and they seldom (if ever) get warm on the bottom. I normally just do modest use - web browsing, e-mail, word processor, etc. However, I just picked up a 2011 11" MBA (base model) at Best Buy (a steal at $740). I've noticed that under the same use, the base of the MBA gets noticeably warm. The CPU temps ton't seem to be all that different measuring with Temperature Monitor, but the heat is a bit uncomfortable working with the MBA on my lap. Reminds me of the old (2008-2010) MBAs. Anyone have the same experience? Is the difference in case design and/or the C2D vs. I5 cpus? Could it have to do with the integrated graphics on the I5 vs. the separate gnu on the 2010 MBA?

Thanks!

 

[dmelgar]

I'm wondering this too. I have a 2010 11" MBA that I'm debating upgrading to a 2012 11" MBA. I've bought the 2012 but haven't decided about keeping it.

My 2010 almost always runs cool and quiet.

The 2012 has sometimes been running hot even though CPU usage is very low.

I've run some tests checking max power drain. Having display cranked up and CPU pegged, I can get the 2010 11" to use 20.1 watts. The 2012 in the same circumstance is using 31.2 watts.

Thats a huge difference and kills battery life. While those tests were running from a full battery, my 2 yr old MBA said 1hr:30min remaining, while the 2012 said about 1hr.

I'm not too keen to spend a ton of money to get 30% less battery life. I haven't yet tested my more typical usage on battery.

 

[Barna Biro]

1:30 hours and 1 hour? So you're for sure running some intensive processes there... did you ever consider that although the newer model has less time remaining it most likely executed a lot more things in that entire time?

Seriously... You can't just say: "Awww, my Lambo's engine heats up quite badly compared to my old Prius". Try to be objective and don't just blindly look at minutes... If the new model would heat up more and eat more energy AND EXECUTE THE EXACT SAME NUMBER OF OPERATIONS IN THE SAME TIME ( or very close time-wise ) as the old one AND also have less battery remaining at the end... then YES, we'd have a problem!

I seriously doubt that's the case...

 

[dmelgar]

1:30 hours and 1 hour? So you're for sure running some intensive processes there... did you ever consider that although the newer model has less time remaining it most likely executed a lot more things in that entire time?

Seriously... You can't just say: "Awww, my Lambo's engine heats up quite badly compared to my old Prius". Try to be objective and don't just blindly look at minutes... If the new model would heat up more and eat more energy AND EXECUTE THE EXACT SAME NUMBER OF OPERATIONS IN THE SAME TIME ( or very close time-wise ) as the old one AND also have less battery remaining at the end... then YES, we'd have a problem!

I seriously doubt that's the case...

Please re-read the post. I stated two different cases.
In one situation with low CPU usage (basically nothing running), the 2012 is running hotter than the 2010.

In the other case I was purposely pegging the CPU as a test. "Pegged" means maxed out the CPU so its 100% busy. You can easily peg a CPU by running a runaway process that does a busy loop rather than waiting for something to do. Many applications do this by accident, some do it on purpose because they're lazy. Its the pegged CPU testcase where battery life dropped from 1:30 on the 2010 to 1hr on the 2012.

 

[Barna Biro]

You are seemingly not getting my point... You are just looking at minutes without considering that during the test, in the same amount of time, the new model might have executed many more operations than the old one ( thus getting the same job done faster, but of course draining the battery faster ).

I'm sorry that I don't believe your descriptions... but your measurements are far from accurate, you have no proper comparison data, you don't know how many cycles did the 2012 run compared to the 2012 model and so on and so forth. All I am saying is that you're doing the entire comparison wrong and with your affirmations, you are unwillingly / unknowingly ( hopefully ) misleading many of the topic readers.

For example: Having a Lamborghini motor running at full speed and a Prius motor at full speed will of course leave the Prius running long after the Lamborghini is out of fuel... but at that time the Lamborghini would have already traveled 2-3 times the distance ( if not more ) than the Prius... so although the Prius will run for a longer period of time, it will still spend the remaining time trying to catch up with the Lambo's position.

The above might not be the best way of putting it, but hopefully you'll now understand what I meant.

 

[GGJstudios]

Does anyone have any data on average temps for the MBA models of the last three iterations?

There's no such thing as "average temps", since there are a great number of factors that contribute to temperatures, including ambient temp, apps/widgets/processes running, computer configuration and settings, etc.

I normally just do modest use - web browsing, e-mail, word processor, etc. However, I just picked up a 2011 11" MBA (base model) at Best Buy (a steal at $740). I've noticed that under the same use, the base of the MBA gets noticeably warm. The CPU temps ton't seem to be all that different measuring with Temperature Monitor, but the heat is a bit uncomfortable working with the MBA on my lap.

Going by "feel" is very inaccurate. Also, as Apple states in the MacBook Air User Guide:

If your MacBook Air is on your lap and gets uncomfortably warm, remove it from your lap and place it on a stable work surface.

 

[TLewis]

You are seemingly not getting my point... You are just looking at minutes without considering that during the test, in the same amount of time, the new model might have executed many more operations than the old one ( thus getting the same job done faster, but of course draining the battery faster ).

While this is true, many people do not care about this. For example, in the case of games, people care about:

1. Does it work/play well? This item is irrelevant to this thread, and so I'm going to assume that the answer is, "yes".
2. How long can people play on battery power?
3. How hot does the MBA get?

Note that operations-per-unit-of-time is not really important here (except to maximize item #1).

Of course, operations-per-unit-of-time do matter in the case of scientific or engineering computations (or photoshop transformations or ripping, etc., etc.)

 

[jsolares]

You are seemingly not getting my point... You are just looking at minutes without considering that during the test, in the same amount of time, the new model might have executed many more operations than the old one ( thus getting the same job done faster, but of course draining the battery faster ).

I'm sorry that I don't believe your descriptions... but your measurements are far from accurate, you have no proper comparison data, you don't know how many cycles did the 2012 run compared to the 2012 model and so on and so forth. All I am saying is that you're doing the entire comparison wrong and with your affirmations, you are unwillingly / unknowingly ( hopefully ) misleading many of the topic readers.

For example: Having a Lamborghini motor running at full speed and a Prius motor at full speed will of course leave the Prius running long after the Lamborghini is out of fuel... but at that time the Lamborghini would have already traveled 2-3 times the distance ( if not more ) than the Prius... so although the Prius will run for a longer period of time, it will still spend the remaining time trying to catch up with the Lambo's position.

The above might not be the best way of putting it, but hopefully you'll now understand what I meant.

If you have a badly programmed process you won't get more out of less battery, just less battery.

Not everyone needs the increase in cpu performance vs the 2010 model. and the increase in heat might not be acceptable.

He doesn't care how much more the 2012 actually accomplished, which is nothing at all since he was probably just running several `yes > /dev/null`

 

[calvol]

I have a 2010 13-UMBA and picked up a friend's 2011 13-UMBA-i7 and first thing I noticed was the heat spot on my left leg, I thought, "wow, I've never felt that before!". What's causing it is the Sandy/Ivy Bridge TurboBoost and under-powered GPU. TB kicks in more often to make up for the weaker GPU, causing a "peaky" heat profile, especially when a lot of windows are open, thus taxing the GPU. I am perfectly content with my 2010 for that reason-- runs cooler and has longer batt life.

 

[AtmChm]

I have a 2010 13-UMBA and picked up a friend's 2011 13-UMBA-i7 and first thing I noticed was the heat spot on my left leg, I thought, "wow, I've never felt that before!". What's causing it is the Sandy/Ivy Bridge TurboBoost and under-powered GPU. TB kicks in more often to make up for the weaker GPU, causing a "peaky" heat profile, especially when a lot of windows are open, thus taxing the GPU. I am perfectly content with my 2010 for that reason-- runs cooler and has longer batt life.

Thanks, that's what I was wondering. Any experience with the 2012 MBA? Is the heat issue the same?

 

[calvol]

Thanks, that's what I was wondering. Any experience with the 2012 MBA? Is the heat issue the same?

I've seen some threads that indicate the 2012 runs even hotter due to the smaller die size and tri-gate transistor. However, the GPU is better, so perhaps TB won't kick in as much.

 

[Barna Biro]

All I was saying that his affirmation of the 2010 MBA having 1:30 hours left while the 2012 MBA has only 1 hour left is not accurate and misleading. His comparison is off and he is completely ignoring the fact that the 2012 might have eaten through battery faster because it processed a lot more in that time frame than the 2010 model did.

Which part are you people failing to understand?

If he doesn't need the processing speed, then of course he has no need for the new model.