2012 MBAs: Battery and heat difference between Ivy Bridge i5 and i7

[Jobsian]

In the absence of any comprehensive battery and heat benchmarks so far, I was wondering if anyone had any ideas/thoughts on the question.

Last year's MBA Sandy Bridge i5 and i7 models showed little difference in battery life according to Anandtech, though I can't remember the exact margin. However in terms of heat there were a lot of anecdotes on this forum that the i7's were running hot on idle compared with the i5's (a major turn-off for me).

Now given this, normally I wouldn't hesitate to get the i5, except for the fact that early benchmarks are showing an even greater performance difference between 2012 11" model i5 Vs 2012 11" model i7, around 25%. For some reason, the geekbench difference between the two 2012 13" models is minuscule (anomalous result?).

I'm aware that this won't translate into a noticeable difference for everyday performance but I am tempted by this power difference for those times where I will stress the machine (eg video encoding).

But it depends on heat and battery differences. I've seen conflicting posts on this around the place already and I'm not sure of the technicalities. Intel have confirmed that both the ivy i5 and i7 will run at 14W, but is that necessarily an indication of identical power draw and battery life? And what about heat, anyone know the technical differences as to why the i7s were hotter idle last year and are those differences still present this year?

[kodeman53]

The short answer is that no one knows until someone does tests. The people on here who claim the I7 uses more power and creates more heat are simply guessing that a faster CPU always does this.

[Mjmar]

I am also very curious about this. The ivy bridge i5 is faster than last year's i7 so if it is cooler and quieter it seems like the better choice for most users.

[jgbr]

But you could look at this way

the i7 gets a task done in 40 seconds at full power and heat

the i5 gets the same task done in 60 seconds at full power and heat

The same amount of energy has been used for either, although one pulled more , it did it more quickly. So battery power must be no different.

Heat on the other hand, id rather have it hot for 40 seconds then 60 seconds if you were protecting the longevity of your kit.

[Jobsian]

Originally Posted by Jobsian

Now given this, normally I wouldn't hesitate to get the i5, except for the fact that early benchmarks are showing an even greater performance difference between 2012 11" model i5 Vs 2012 11" model i7, around 25%. For some reason, the geekbench difference between the two 2012 13" models is minuscule (anomalous result?).

Looks like it was a mistake after all, following posted in the news forum:

Originally Posted by rot@ti.org

The Macbook Air chart is mislabeled. The number for the 1.8 ghz 13-inch model says 6,915. But the original chart from PrimateLabs says 6,197. The relative size of the bar is closer to 6,197 than 6,915.

[PeterJP]

Originally Posted by jgbr

the i7 gets a task done in 40 seconds at full power and heat

the i5 gets the same task done in 60 seconds at full power and heat

The same amount of energy has been used for either, although one pulled more , it did it more quickly. So battery power must be no different.

Heat on the other hand, id rather have it hot for 40 seconds then 60 seconds if you were protecting the longevity of your kit.

That would be one way to look at it. I think your estimates of battery life are quite accurate, looking at the benchmarks of the previous MBA i5 vs i7. The i7's battery lasted only a little bit shorter.

What I disagree with, though, is that "short hot heat" bit. What I read in this forum was that several i7 users reported their notebook was quite warm even in idle, whereas the i5 users didn't. Even the slightest hint at this possibility is enough for me to run away screaming. I had one notebook that did this. Heat problems never get better, they degrade over time. I didn't like opening my notebook on a regular basis (say, once a year after the first 2-3 years) to clean up the thermal paste and to replace it with new paste. For that reason, I will never go for a notebook that has reports of being "fairly warm in idle". I don't really need the extra processing power (though it would be nice when I start organising my photo collection shot in RAW). I prefer a laptop that works stably for 5 years or more.

Just my $0.02, of course.


Peter.

[soldierblue]

It's pretty widely reported that desktop Ivy chips run quite a bit hotter than Sandy chips, especially when overclocked. No idea if this translates to the mobile variants.

[Jobsian]

Originally Posted by PeterJP

That would be one way to look at it. I think your estimates of battery life are quite accurate, looking at the benchmarks of the previous MBA i5 vs i7. The i7's battery lasted only a little bit shorter.

What I disagree with, though, is that "short hot heat" bit. What I read in this forum was that several i7 users reported their notebook was quite warm even in idle, whereas the i5 users didn't. Even the slightest hint at this possibility is enough for me to run away screaming. I had one notebook that did this. Heat problems never get better, they degrade over time. I didn't like opening my notebook on a regular basis (say, once a year after the first 2-3 years) to clean up the thermal paste and to replace it with new paste. For that reason, I will never go for a notebook that has reports of being "fairly warm in idle". I don't really need the extra processing power (though it would be nice when I start organising my photo collection shot in RAW). I prefer a laptop that works stably for 5 years or more.

Just my $0.02, of course.


Peter.

Likewise, if there is any idle heat issues I'm steering clear of i7. I don't mind a bit of heat when I'm stressing the thing, but not when I'm reading email.

[cnixon]

You guys are seriously making me doubt my choice to purchase the i7 11 inch. Should I consider changing my order?

[rygamble]

Originally Posted by cnixon

You guys are seriously making me doubt my choice to purchase the i7 11 inch. Should I consider changing my order?

If you had gotten a 13", yes (as there is only a very small improvement between the two). However, the 11" is showing upwards of 25% improvement between i5 and i7. I say just get it, and if it feels like it's running too warm for your liking then just take it back.

[Jobsian]

Originally Posted by rygamble

If you had gotten a 13", yes (as there is only a very small improvement between the two). However, the 11" is showing upwards of 25% improvement between i5 and i7. I say just get it, and if it feels like it's running too warm for your liking then just take it back.

Don't forget that the performance difference between the two processors in the 13" model isn't as small as first thought due to a mistake (see my post above).

[joudbren]

I just picked up a 2.0ghz i7 11" MacBook Air tonight at the Apple store. I'm just surfing the web right now and it's been on for a couple of hours already. The unit is barely warm so far and I don't think I've heard the fan yet. Apparently fan is redesigned with some new blade technology that is uber quiet.

Mighty impressed so far but I can tell you at least for surfing the web, let alone idling, the heat is imperceptible at this point. Haven't tried anything hard core yet though. Cheers!

James

[Jobsian]

Originally Posted by joudbren

I just picked up a 2.0ghz i7 11" MacBook Air tonight at the Apple store. I'm just surfing the web right now and it's been on for a couple of hours already. The unit is barely warm so far and I don't think I've heard the fan yet. Apparently fan is redesigned with some new blade technology that is uber quiet.

Mighty impressed so far but I can tell you at least for surfing the web, let alone idling, the heat is imperceptible at this point. Haven't tried anything hard core yet though. Cheers!

James

Excellent, thank you, an update of how long the battery approximately lasted for you would be great, as would any other anecdotes.

Eagerly awaiting some formal benchmarks too. So far the anecdotes for the i7 are more positive than last year, don't want to make my decision just yet though in case the i5 somehow eeks out significantly more battery

[joudbren]

Originally Posted by Jobsian

Excellent, thank you, an update of how long the battery approximately lasted for you would be great, as would any other anecdotes.

Eagerly awaiting some formal benchmarks too. So far the anecdotes for the i7 are more positive than last year, don't want to make my decision just yet though in case the i5 somehow eeks out significantly more battery

Can't tell you anything on the battery yet as it will have to go through a few charge cycles first to allow for its maximum capacity. No benchmarks yet other than the SSD. Doing around 410MB/s on writes and 450MB/s on reads. (256GB Samsung SSD) Absolute greased lightening! Battery life "seems" similar to my 2011 11" Air but again, too early to tell until the battery gets conditioned. Cheers!

James

[ericv232]

Originally Posted by joudbren

Can't tell you anything on the battery yet as it will have to go through a few charge cycles first to allow for its maximum capacity. No benchmarks yet other than the SSD. Doing around 410MB/s on writes and 450MB/s on reads. (256GB Samsung SSD) Absolute greased lightening! Battery life "seems" similar to my 2011 11" Air but again, too early to tell until the battery gets conditioned. Cheers!

James

Hey James,
Does the keyboard between the 2011 11" Air seem similar or different to the keyboard on the 2012 11" Air?

[joudbren]

Originally Posted by ericv232

Hey James,
Does the keyboard between the 2011 11" Air seem similar or different to the keyboard on the 2012 11" Air?

Pretty much virtually identical from what I can tell. Initially I thought the keys might be just a teensy bit stiffer on the new 2012 but it is so slight it can probably be attributed to it just being newer and not broken in yet. Otherwise, they're identical layout, colour, back-lighting, etc.

J.

[Zzari]

Originally Posted by joudbren

Can't tell you anything on the battery yet as it will have to go through a few charge cycles first to allow for its maximum capacity. No benchmarks yet other than the SSD. Doing around 410MB/s on writes and 450MB/s on reads. (256GB Samsung SSD) Absolute greased lightening! Battery life "seems" similar to my 2011 11" Air but again, too early to tell until the battery gets conditioned. Cheers!

James

Apple ships the batteries pre-calibrated...it's unlikely that "conditioning" will lead to any improvement. I was hoping that Ivy Bridge would increase 11" battery, but from reading around the last two days it seems that battery life is pretty much the same (maybe an extra 15-20 minutes from Ivy on average, but nothing spectacular).

[ericv232]

Originally Posted by joudbren

Pretty much virtually identical from what I can tell. Initially I thought the keys might be just a teensy bit stiffer on the new 2012 but it is so slight it can probably be attributed to it just being newer and not broken in yet. Otherwise, they're identical layout, colour, back-lighting, etc.

J.

Good to know. Thanks.

[Tumej]

is MBA really so quiet? no fan making uber noices?
Ny Toshiba laptop is awful, when i turn it on the fan sounds like hoover...

[amityngmedia]

Is it true that MBA fan does make noise that a great advantage.As i'm having HP G42 it makes a lot of noise and having an i5 processor it runs very slow.

[theSeb]

I would suggest waiting for full reviews of production machines, but you can get a good idea from this preview on Anandtech's website about mobile dual core Ivy Bridge

All indications right now are that Ivy Bridge is far more important for Intel’s laptop division than for their desktops. Yes, we'll eventually get dual-core Ivy Bridge on the desktop as well, but many of those chips will be saddled by less than impressive HD 2500 graphics, which Anand discusses in his i5-3470 review. So why would Intel put all that die space and energy into their IGP if they’re just going to disable half of it on most desktops? Because on desktops, integrated graphics performance isn’t particularly important; you can always add a discrete GPU (unless you’ve got a really small/proprietary system). On laptops, a large percentage ship without any sort of dedicated GPU, and it’s often a bottleneck for home users.

So just what does Ivy Bridge bring to the party that you couldn’t get with Sandy Bridge Ultrabooks? In a word: more. More CPU performance—the i5-3427U we tested today is typically close to i5-2410M performance, and often 20% faster—or more—compared to Sandy Bridge Core i7 Ultrabooks. More GPU performance: HD 4000 in IVB ULV is generally faster than HD 3000 in SNB standard voltage CPUs. And you get all that with similar or slightly better battery life. You also get less: far less bulk and weight to carry around. We’re basically looking at the performance of a laptop that used to weigh six pounds in a three pound chassis. If you’re someone who carries their laptop around a lot, an Ultrabook would make for an excellent companion—whether for business, school, or some other use. They’re light, fast enough, and get great battery life, and they’re small enough to fit in a purse or a small laptop bag—no more giant laptop carry ons, thank you very much!

http://www.anandtech.com/show/5872/i...abook-review/7

[joudbren]

Originally Posted by Zzari

Apple ships the batteries pre-calibrated...it's unlikely that "conditioning" will lead to any improvement. I was hoping that Ivy Bridge would increase 11" battery, but from reading around the last two days it seems that battery life is pretty much the same (maybe an extra 15-20 minutes from Ivy on average, but nothing spectacular).

That hasn't been my experience including my 2011 MBA from last year and every phone I've owned in the last 5+ years. Lithium technology batteries are widely known to require several charge cycles before they reach their maximum potential. It might not make a big difference in runtime but regardless, I still don't count on any battery readings from brand new cells until they've gone through a few cycles. Cheers!

J.

[Jobsian]

Utterly dying for the Anandtech benchmarks. It seems to be good news so far anecdotally, but would love to see confirmation.

I'm tempted to go to an Apple store and benchmark one of these things myself for 5 hours

[palpatine]

This is anecdotal and far from a rigorously tested use, but I got 4:47 out of the 11" MBA (see the 11" MBA thread for usage details). That's not great, but it isn't bad either.

[Jobsian]

Originally Posted by palpatine

This is anecdotal and far from a rigorously tested use, but I got 4:47 out of the 11" MBA (see the 11" MBA thread for usage details). That's not great, but it isn't bad either.

Thanks, that's getting close to swaying me.