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

GoPro

macrumors member
Original poster
Dec 10, 2012
82
0
Is it me or are the new macbook air's slower than the last generation?

i.e. 1.8GHz (up to 2.8GHz with turbo boost) of 2012 vs the 1.3GHz (2.4GHz with turbo boost) of the 2013 models.

Could it be that the macbook air's were getting too powerful, so powerful that they were matching macbook pro's for performance?

For example, my i7 2.0GHz outperforms the base macbook pro retina 13" (with turbo mode on both models active) and is on par with the higher spec one (again, with turbo mode active on both models).

Did apple reduce the speed of these processor's to save battery life and market it as an ideal laptop for somebody who is constantly on-the-go?
 

technowar

macrumors 6502
Apr 1, 2011
371
1
Cebu, Philippines
You can't compare both processors with its clock frequency alone. 2012 is an Ivy Bridge [tick] processor and is one year old. 2013 on the other hand uses Haswell [tock] processor. Tick is shrinking of process technology of the previous microarchitecture, while Tock is a new microarchitecture. This would mean another level higher than the previous one.
 

TechZeke

macrumors 68020
Jul 29, 2012
2,454
2,286
Dallas, TX
There is more to processor speed than just clock speed. With a more efficient processor, you can do more things with less energy. There also other variables such as the cache, architecture, etc.

If clock speed was the only thing that determined performance, we would have had 9Ghz processors years ago.
 

Uliman

macrumors member
Mar 23, 2010
71
2
I am interested in the answer to this question as well. I have the 2012 2.0 I7 with 8gb RAM and was looking to upgrade but after the very disappointing WWDC offerings I probably will have to wait until the Retina MBP gets slimmed down and gets Haswell.

Will the i7 1.7 Haswell generally be faster than my i7 2.0 Ivy? Looking forward to tests of course that will give real world results.
 

GoPro

macrumors member
Original poster
Dec 10, 2012
82
0
There is more to processor speed than just clock speed. With a more efficient processor, you can do more things with less energy. There also other variables such as the cache, architecture, etc.

If clock speed was the only thing that determined performance, we would have had 9Ghz processors years ago.

Normally I would agree. However up until now the largest decrease in clock speed has been the transition from late 2010 models (1.83GHz) to 2011 models (1.6GHz) which is a difference of less than 2Ghz. This time it's a huge 5ghz difference in clock speed.

IMO apple have limited the processor's so that they offer a great battery life, at the expense of performance. OR, the performance is the same (not inferior) while offering much greater battery life.
 

Mr. RPG

macrumors 6502a
Sep 21, 2012
566
0
Since the Airs will ship immediately we should know by the end of the week if they are indeed faster through benchmarks.
 

mattferg

macrumors 6502
May 27, 2013
380
22
There are some serious misunderstandings of how a processor works going on in this thread.

For a start, turbo boost means that even if the BASE clock speed is lower, the current processors can just amp up the clock speed to match the 2012 models and be as powerful that way. Kinda how turbo boost works, and means judging a processor by it's base clock speed isn't really the way to go any more.

Secondly, much like how an i5 outperforms a Core 2 Duo at the same clock speeds even without HT, the Has well processor is faster than the Ivy at matched clock speeds, and even still faster at lower ones.

Another factor is the graphics boost - HD5000 is far more powerful than 4000.

So yes, the new MacBook Air is more powerful than the last one. To believe anything else is just ignorant.
 

mattferg

macrumors 6502
May 27, 2013
380
22
Is it me or are the new macbook air's slower than the last generation?

i.e. 1.8GHz (up to 2.8GHz with turbo boost) of 2012 vs the 1.3GHz (2.4GHz with turbo boost) of the 2013 models.

Could it be that the macbook air's were getting too powerful, so powerful that they were matching macbook pro's for performance?

For example, my i7 2.0GHz outperforms the base macbook pro retina 13" (with turbo mode on both models active) and is on par with the higher spec one (again, with turbo mode active on both models).

Did apple reduce the speed of these processor's to save battery life and market it as an ideal laptop for somebody who is constantly on-the-go?

Your figures are incorrect.

The 2012 model i5 was 1.7 - 2.6 and the 2013 is 1.3 - 2.6. Just as much availabke power and speed. Same goes for the i7, where the 2013 model actually has more power available. 2012 was 2 - 3.2 and 2013 is 1.7 - 3.3.

----------

It appears that in raw Benchmarking - the 2013 Airs are slower...

Averaging about 6000 so far in Geekbench averages.

http://browser.primatelabs.com/geekbench2/search?utf8=✓&q=4250U

Versus the 2012 - the 1.8Ghz was ~6100 and upwards of 7600 for the 2.0 i7 models:

http://browser.primatelabs.com/geekbench2/search?utf8=✓&q=macbookair5,2

Nope

http://browser.primatelabs.com/geekbench2/2030196
 

jmcube

macrumors member
Jun 6, 2013
58
10
You need to be careful between 64/32, it seems.

It is ~6900 for the Intel Core i7-3667U... 2000 running in 32bit mode.
Intel Core i5-3427U got only ~6000 in 32bit mode.

It means that the i5 are equivalent (but the haswell has better battery and better graphics).

I actually think that if anything the haswell i7 will be better (as it doesn't lose as much Hz) and has higher max turbo.

Jon
 

mattferg

macrumors 6502
May 27, 2013
380
22
You need to be careful between 64/32, it seems.

It is ~6900 for the Intel Core i7-3667U... 2000 running in 32bit mode.
Intel Core i5-3427U got only ~6000 in 32bit mode.

It means that the i5 are equivalent (but the haswell has better battery and better graphics).

I actually think that if anything the haswell i7 will be better (as it doesn't lose as much Hz).

Jon

Completely agree, and the Haswell i7 has an extra 100mhz at the top end.
 

torana355

macrumors 68040
Dec 8, 2009
3,609
2,676
Sydney, Australia
The new Airs have already been bench marked and are indeed slower the the 2012 models. Haswell would bring a small bump in performance but not enough to make up for the loss of 400mhz clockspeed. Im not fussed as i use my MBA for very light tasks anyway and prefer more battery life.
 

LeeM

macrumors 6502a
Jan 1, 2012
603
0
IMO apple have limited the processor's so that they offer a great battery life, at the expense of performance. OR, the performance is the same (not inferior) while offering much greater battery life.

i agree. i think its more to differentiate the air from the pro retina. with the retina being so thin it makes the air sort of pointless, unless the air is less powerful and has a longer battery life and is cheaper. the air becomes the go to machine if you want to use word/pages while youre travelling and the pro is there for more intensive tasks
 

Mrbobb

macrumors 603
Aug 27, 2012
5,009
209
Sounds like at least some of power savings tricks are software. Am wondering how Mavericks will improve, or not, to an old hardware.
 

mattferg

macrumors 6502
May 27, 2013
380
22
The new Airs have already been bench marked and are indeed slower the the 2012 models. Haswell would bring a small bump in performance but not enough to make up for the loss of 400mhz clockspeed. Im not fussed as i use my MBA for very light tasks anyway and prefer more battery life.

There hasn't been a LOSS of 400mhz clockspeed and the benchmarks are pretty much the same.
 

whtrbt7

macrumors 65816
Jun 8, 2011
1,015
73
Perceived speed is different from clock speed and processor bench speed. One of the bottlenecks to speed for the 2012 MBAs was the graphics processing. Haswell SoCs are designed to deliver better graphics performance with lower power consumption as well as a lot of other features. Perceived speed of the 2013 MBAs should be about 10-15% faster than a comparable 2012 model running side by side.
 

snapdragonx

macrumors regular
Oct 31, 2012
141
1
It appears that in raw Benchmarking - the 2013 Airs are slower...

Averaging about 6000 so far in Geekbench averages.

http://browser.primatelabs.com/geekbench2/search?utf8=✓&q=4250U

Versus the 2012 - the 1.8Ghz was ~6100 and upwards of 7600 for the 2.0 i7 models:

http://browser.primatelabs.com/geekbench2/search?utf8=✓&q=macbookair5,2

Let's assume this is correct. The difference would be insignificant.

Plus, take into account the extra 4-5 hours battery life, and the fact that Engadget just reported the SSD read speed at 700MB/s+, plus the boost in graphics performance

That's a big bonus, and worth the sacrifice of 100 or even 1000 geekbench points.

Then take into account that benchmarks don't always relate to real-world performance. The perceived performance will no doubt be improved.
 

paulrbeers

macrumors 68040
Dec 17, 2009
3,963
123
Normally I would agree. However up until now the largest decrease in clock speed has been the transition from late 2010 models (1.83GHz) to 2011 models (1.6GHz) which is a difference of less than 2Ghz. This time it's a huge 5ghz difference in clock speed.

IMO apple have limited the processor's so that they offer a great battery life, at the expense of performance. OR, the performance is the same (not inferior) while offering much greater battery life.

Actually it isn't apple at all on this one. The fastest Haswell ULV processor is the 1.7ghz. Kinda hard to include faster processors WHEN THEY DON'T EXIST!!!
 

fyrefly

macrumors 6502a
Jun 27, 2004
614
48

Why are you quoting a Windows 7 benchmark, exactly?

Let's assume this is correct. The difference would be insignificant.

Plus, take into account the extra 4-5 hours battery life, and the fact that Engadget just reported the SSD read speed at 700MB/s+, plus the boost in graphics performance

That's a big bonus, and worth the sacrifice of 100 or even 1000 geekbench points.

Then take into account that benchmarks don't always relate to real-world performance. The perceived performance will no doubt be improved.

Agreed. I was just posting Raw benchmark numbers - and you're 100% right that those don't always directly translate. The improved Graphics, battery life and SSD speed increases will certainly delight many.

But even if equal - the Haswell advantage comes mostly in battery and graphics, at the expense of any gains in raw power. This is not the jump from Core2 to Core i like the 2010-2011 MBAs, or even the Ivy->Sandy jump in 2012 which resulted in jumps in Raw power as well as Graphics benchmarks.

Apple has clearly positioned the Air as a portable road warrior machine, probably to differentiate it from the MBP as the "power portable".

My bottom line take (total IMHO) is unless you want to game on the Intel HD 5000, or need 8+ hours battery life (let's face it the current Air's don't get 7, so the 2013 Airs won't get 12 actual hours) you're probably fine with a 2011 or 2012 MBA vs. getting a Haswell machine.
 

MacLappy

macrumors 6502a
Jul 28, 2011
529
390
Singapore
Is it me or are the new macbook air's slower than the last generation?

My i7 2.0GHz outperforms the base macbook pro retina 13" (with turbo mode on both models active) and is on par with the higher spec one (again, with turbo mode active on both models).

Did apple reduce the speed of these processor's to save battery life and market it as an ideal laptop for somebody who is constantly on-the-go?

I am interested in the answer to this question as well. I have the 2012 2.0 I7 with 8gb RAM and was looking to upgrade but after the very disappointing WWDC offerings I probably will have to wait until the Retina MBP gets slimmed down and gets Haswell.

Will the i7 1.7 Haswell generally be faster than my i7 2.0 Ivy? Looking forward to tests of course that will give real world results.

Am in the same situation as you guys.

Talking about just CPU speed, clock for clock haswell is faster. Unfortunately this time around we are not seeing a direct increase in clock speed like the last update(sandybridge to ivybridge). Here are some of the know results.

Geekbench 2 Results

MacBook Air 2013 i5 4250U = 6043
MacBook Air 11" 2012 i5 = 5758
MacBook Air 13" 2012 i5 = 6120
MacBook Air 2012 i7 = 7024

On the processor front
- slight upgrade going from 11" 2012 to 2013 base processor
- very slight downgrade from 13" 2012 to 2013 base processor
- noticeable downgrade from i7 2012 to 2013 base processor
- unconfirmed but most likely results i7 2012 to i7 2013 should be a slight upgrade

The main focus seems to be on battery life this year. Last year was probably a really special case, seeing how it was a tick plus. Yeah a nice jump in both cpu/gpu performance in one generation is pretty awesome. core2duo to sandy was a huge leap for the cpu as it was a multi-generation leap. Coming back to the topic of i7 3667u vs i7 4650u, from the i5 results we can guesstimate i7 4650u would probably be slightly faster, my guess less than 5%. I might skip this year update as battery life is not a major concern for me, the 40% increase in gpu isn't really going to change the way i use my MacBook Air. I will still have to leave the games i play at the same settings, the jump is not noticeable enough for anything drastic, remember depreciating returns are in order here. Going from 8fps to 22fps on benchmark is not the same as going from 22fps to 29fps. 8fps to 22 fps changes the whole playability of the game, the next noticeable jump would be from 22fps to something like 60fps, at least that's how I understand it. Furthermore, while I do enjoy some more eye candy, the greater need is more cpu bound for me, like 25 man raid in WOW, 4 way multiplayer in SC2 and multiplayer in D3. The bottleneck in these scenarios are mostly CPU base, which is completely not addressed in this update. Hopefully next year we see a good increase in both cpu/gpu along with the already awesome battery life from this year.
 

MacLappy

macrumors 6502a
Jul 28, 2011
529
390
Singapore
My jump from Sandybridge to Ivybridge, benchmark results.
 

Attachments

  • Screen Shot 2013-05-21 at 5.13.40 PM.png
    Screen Shot 2013-05-21 at 5.13.40 PM.png
    145.2 KB · Views: 144
  • Screen Shot 2013-06-09 at 10.51.15 PM.png
    Screen Shot 2013-06-09 at 10.51.15 PM.png
    98.6 KB · Views: 136

ayres

macrumors 6502
Sep 27, 2010
289
50
i agree. i think its more to differentiate the air from the pro retina. with the retina being so thin it makes the air sort of pointless, unless the air is less powerful and has a longer battery life and is cheaper. the air becomes the go to machine if you want to use word/pages while youre travelling and the pro is there for more intensive tasks

x2

though it's certainly good for more than just word/pages

----------

Perceived speed is different from clock speed and processor bench speed. One of the bottlenecks to speed for the 2012 MBAs was the graphics processing. Haswell SoCs are designed to deliver better graphics performance with lower power consumption as well as a lot of other features. Perceived speed of the 2013 MBAs should be about 10-15% faster than a comparable 2012 model running side by side.

not to mention, any improvement in the read/write as was highlighted in the keynote and on the apple's website. i'm curious to learn more about this in the coming weeks.
 

GoPro

macrumors member
Original poster
Dec 10, 2012
82
0
Your figures are incorrect.

The 2012 model i5 was 1.7 - 2.6 and the 2013 is 1.3 - 2.6. Just as much availabke power and speed. Same goes for the i7, where the 2013 model actually has more power available. 2012 was 2 - 3.2 and 2013 is 1.7 - 3.3.

nope. the 2012's were 1.8GHz.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.