The Optional 1.8GHz Core i7

In the usual Apply fashion, in order to get the faster CPU options you have to buy the more expensive default configuration. That means the $999 11-inch MacBook Air can't be upgraded to a faster CPU at order, while the $1199 version can. The same goes for the $1299 13-inch model, you need the $1599 configuration for that.

If you're buying the 11, I'd say it's worth while to get the 1.8GHz option. There's a big enough performance difference between the base 13 and base 11 that the CPU upgrade makes sense.

I didn't have a 1.8GHz system on hand but luckily I knew someone who did. Mark Rein is VP of Epic Games and happens to live a few minutes from my house. Mark also happened to be driving down to the Apple store on Friday to pick up a fully optioned (the only way Mark rolls) 13-inch MacBook Air with a 1.8GHz Core i7. I hitched a ride with him and borrowed his system for a couple of benchmark runs.

Let's look at the clock speed differences between the parts:

2011 Apple MacBook Air CPU Comparison
1.6GHz Core i5 1.7GHz Core i5 1.8GHz Core i7
Base Clock Speed 1.6GHz 1.7GHz 1.8GHz
Max SC Turbo 2.3GHz 2.7GHz 2.9GHz
Max DC Turbo 2.0GHz 2.4GHz 2.6GHz

Compared to the 1.7GHz chip in the 13, you're looking at around an 8% performance increase in applications that scale perfectly with clock speed. The extra MB of L3 cache is good for a little more performance as well. Compared to the 1.6GHz chip in the 11 however, the gains on paper are pretty substantial. What about in the real world?

2011 Apple MacBook Air CPU Comparison
1.6GHz Core i5 1.7GHz Core i5 1.8GHz Core i7
Cinebench R10 - Single Threaded 3154 3770 4083
Percent Increase with 1.8GHz i7 29.5% 8.3% -

With only a single heavy thread active the i7 is nearly 30% faster than the 1.6GHz Core i5 found in the 11-inch MacBook Air. The advantage over the 1.7GHz Core i5 is only 8.3% however.

2011 Apple MacBook Air CPU Comparison
1.6GHz Core i5 1.7GHz Core i5 1.8GHz Core i7
Cinebench R10 - Multithreaded 6496 7609 8234
Percent Increase with 1.8GHz i7 26.8% 8.2% -

Crank up the threads and the performance advantage drops a bit over the 1.6GHz Core i5 but we're still talking about a nearly 27% advantage here. The i7 is still only around 8% faster than the 1.7GHz part.

2011 Apple MacBook Air CPU Comparison
1.6GHz Core i5 1.7GHz Core i5 1.8GHz Core i7
Cinebench 11.5 1.93 2.20 2.46
Percent Increase with 1.8GHz i7 27.5% 11.8% -

Cinebench R11.5 shows the biggest gain over the 1.7GHz Core i5 at 11.8%. Heavier workloads will make better use of the extra cache on the i7.

2011 Apple MacBook Air CPU Comparison
1.6GHz Core i5 1.7GHz Core i5 1.8GHz Core i7
iPhoto Import 38.3 44.4 47.7
Percent Increase with 1.8GHz i7 24.5% 7.4% -

Overall the performance gains aren't super huge over the 13's 1.7GHz Core i5, however if you care about performance the upgrade is worth it for the 11-inch MacBook Air. Unfortunately I didn't have an upgraded 11-inch system on hand to get a good idea of the impact on battery life. In idle workloads I wouldn't expect a huge drop in battery life, but under load we could see the 11 take a hit.

CPU Performance: A Huge Upgrade GPU Performance: A Step Back
Comments Locked

103 Comments

View All Comments

  • netmask254 - Thursday, July 28, 2011 - link

    It's really surprising to me that the huge random speed difference, and how can Apple tolerate such an inconsistent behavior among different machines? I heard that the buyer will randomly get a Sumsung or Toshiba SSD even for the same model, that's too bad. However, maybe most Apple users don't care about it or don't know it.
  • beginner99 - Thursday, July 28, 2011 - link

    Yeah, especially since you would actually pay an apple premium for knowing what you get. But then these are more priced like windows laptops so they probably had to lower some standards.
    The so called ultrabooks according to rumors might even be more expensive than MBA. IMHO especially the 256 GB version is a bargain (if you get the Samsung ssd). That alone costs you like 400$ if you would manually upgrade a cheap windows laptop.
    Isn't the 4k random read the main thing that prevents stuttering issues ins ssd compared to hdds?
  • lyeoh - Thursday, July 28, 2011 - link

    I think the stuttering is more to do with the max access latency.

    You could have a very fast drive that averages 100MB/sec over 1 minute, but if every 30 seconds it "hangs" for one second, you would notice it.

    e.g. for one second your transfer rate drops to 0MB/sec, but for the other 29 seconds it transfers at 104MB/sec.

    That's why many of those benchmarks don't tell you everything about how it feels.

    Same goes for fps you could get 120fps but if what appears on the screen is actually delayed by a 100 milliseconds, it's worse than something that "only" does 60 fps with 5 millisecond delay.

    Not many reviewers test for latency. Anandtech does have some latency tests for a few SSDs, but so far it does not appear to be a standard benchmark.
  • KPOM - Thursday, July 28, 2011 - link

    Interestingly, on the MacRumors forum someone posted results of another benchmark that suggested that the Toshiba outperformed the Samsung on random reads and writes, though the Samsung outperform on sequential.

    http://translate.google.com/translate?u=http%3A%2F...

    Anand says the Samsung outperforms on both. Note that this was also the case in the 2010s. Toshibas started shipping first, and Samsungs made their way into the system in January. Apple seems to have carried them over unchanged to the 2011. I wonder what impact supply chain issues (e.g. the tsunami) and legal issues (e.g. Apple's lawsuit against Samsung) have on their purchasing decisions?
  • fericia - Monday, August 15, 2011 - link

    My wife's 13-inch one came with SM256...
  • arthur449 - Thursday, July 28, 2011 - link

    Light Web Browsing Battery Life
    13-inch MacBook Air (Mid 2011) - Core i5 2.7GHz

    This message will self-destruct after reading.
  • Anand Lal Shimpi - Thursday, July 28, 2011 - link

    Fixed :)

    Take care,
    Anand
  • refresh_time - Thursday, July 28, 2011 - link

    Thank you for a very in-depth review.
    I was ordered the i7 13" without needing the bigger ssd, but after reading this, I think i should exchange it for a i5 :D
    All I wanted from i7 was video work on imovie but if there is only about a 10% gain in performance, I think I'm better with i5.. (right?)
    But i read the mba review here (http://www.gadgetreview.com/2011/07/apple-macbook-... and they seem to recommend i7 highly. what do you think?

    I'll use the return money to buy a 1tb hardrive and a nice sleeve
  • iwod - Thursday, July 28, 2011 - link

    I think we need something thinner, lighter, and more size to choose from. Not everyone's home can fit a bloody 27" Monitor.
  • ImSpartacus - Thursday, July 28, 2011 - link

    Apple doesn't do low margin products.

    Even if they offered a 24" IPS panel with all the fixings, it would approach $500. It's hard to explain to a customer why they need to spend twice as much as a bargain bin 24" TN panel.

    There aren't a lot of 1440p 27" monitors on the market today for less than $1000, so Apple doesn't run into any comparison problems. The camera, speakers and connectivity features are just gravy.

Log in

Don't have an account? Sign up now