Ultimate RevC MBA Vs Ultimate RevD 11" MBA - Is the RevD better in all areas?

[Jobsian]

Will someone moving from a Rev C MBA 2.13GHz, SSD to an "Ultimate" 11"MBA with 1.6GHz, 4GB, 128SSD find any downgrade in performance? (Opening summary and introduction)

My early impressions:
I'm personally finding the 11" more tolerant to YouTube in terms of simultaneous windows, frame-skipping etc, possibly a touch snappier opening programs. Higher fps while gaming on osx. Yet to vm it. (Body)

I can't think of a single thing that's slower so far. Even the speakers seem louder (don't know if this has been objectively verified). Specific benchmark comparisons or personal experiences welcome. (Conclusion and closing statement)

 

[WardC]

I own a Rev. C MacBook Air 2.13GHz / 128GB SSD / 2GB RAM, and I am getting my 11.6" 1.6GHz / 4GB RAM / 128GB Flash MacBook Air on Monday morning. I plan to write a comprehensive review comparing the two. I will run a variety of benchmarks and post complete photos showing both and comparing them. I would like to compare how fast the new 11.6" full-spec MacBook Air is compared to the Rev. C. I suspect the my new 11.6" will beat the old 2.13GHz one in just about everything, except maybe video compression and music file processing, processor intensive tasks. I bet the new Flash storage on the new MacBook Air 11.6" will be faster than the SSD on my rev. C MacBook Air. I notice my current Rev. C takes quite awhile to boot and shutdown, and from looks at YouTube videos of the new units, they boot and shutdown almost instantly. I suspect application loading will be faster too. I can't wait to compare them, and I will post everything on my website. Stay tuned late next week.

 

[Bokes]

I just ran a cinebench on my 2.13 13 inch rev C and I got a score of 84- which is what Macworld scored for the 11inch. (But they also scored my 13inch model at 63)- which makes no sense. My best guess is the old 2.13 and new 11 are very close performance wise. Maybe the 4gigs config gives the 11 the edge- but the big difference IMO is the size factor.

oh yeah- my 13inch 2.13 has backlight keys. (nice)

 

[fyrefly]

The older Air firmware forces the processor to throttle a lot due to heat constraints. Plus the 9400m was down-clocked and even at full-clock is only 1/2 as powerful as the 320m.

Combined with the slower SSD's on the RevC "ultimate" models, that makes the older airs (usually running at 1.6 or 1.86Ghz, rather than the full 2.13Ghz due to heat) performance-wise the same as the newer 1.6Ghz airs with their speedier graphics and SSDs.

IMHO, I had an 11" air, and I still prefer my 13" RevB 1.6Ghz SSD model. Bigger screen, better keyboard, backlit keyboard... if only I could make the SSD faster and the RAM to 4GB, my MBA revB would still be perfect.

 

[freitas]

Here are a couple benchmarks (Prime95 and Xbench). Probably not the fairest comparison as my RevC is undervolted using coolbook and the cpu benchmarks do not run long enough for throttling to kick in. But here are the numbers.

Prime 95
=======
Ultimate 11.6 MBA RevD
Timing 10 iterations at 8192K FFT length. Best time: 279.156 ms.
Timing 10 iterations at 8192K FFT length. Best time: 194.565 ms.
Timing trial factoring of M35000011 with 67 bit length factors. Best time: 10.936 ms.

Ultimate 13.3 MBA RevC
Timing 10 iterations at 8192K FFT length. Best time: 204.937 ms.
Timing 10 iterations at 8192K FFT length. Best time: 117.833 ms.
Timing trial factoring of M35000011 with 67 bit length factors. Best time: 8.176 ms.


Xbench
=======
Ultimate 11.6 MBA RevD
Results 131.18
System Info
Xbench Version 1.3
System Version 10.6.4 (10F3061)
Physical RAM 4096 MB
Model MacBookAir3,1
Drive Type APPLE SSD TS128C
CPU Test 112.43
GCD Loop 187.82 9.90 Mops/sec
Floating Point Basic 89.53 2.13 Gflop/sec
vecLib FFT 74.11 2.44 Gflop/sec
Floating Point Library 178.83 31.14 Mops/sec
Thread Test 188.52
Computation 280.00 5.67 Mops/sec, 4 threads
Lock Contention 142.09 6.11 Mlocks/sec, 4 threads
Memory Test 139.39
System 151.45
Allocate 226.41 831.45 Kalloc/sec
Fill 122.72 5967.09 MB/sec
Copy 138.05 2851.42 MB/sec
Stream 129.11
Copy 121.68 2513.16 MB/sec
Scale 124.32 2568.42 MB/sec
Add 136.72 2912.42 MB/sec
Triad 135.04 2888.93 MB/sec
Quartz Graphics Test 111.63
Line 101.80 6.78 Klines/sec [50% alpha]
Rectangle 115.91 34.61 Krects/sec [50% alpha]
Circle 100.01 8.15 Kcircles/sec [50% alpha]
Bezier 119.93 3.02 Kbeziers/sec [50% alpha]
Text 124.94 7.82 Kchars/sec
OpenGL Graphics Test 94.99
Spinning Squares 94.99 120.50 frames/sec
User Interface Test 120.02
Elements 120.02 550.82 refresh/sec
Disk Test 239.85
Sequential 155.87
Uncached Write 210.09 128.99 MB/sec [4K blocks]
Uncached Write 245.02 138.63 MB/sec [256K blocks]
Uncached Read 78.06 22.84 MB/sec [4K blocks]
Uncached Read 249.37 125.33 MB/sec [256K blocks]
Random 520.06
Uncached Write 312.31 33.06 MB/sec [4K blocks]
Uncached Write 519.97 166.46 MB/sec [256K blocks]
Uncached Read 1422.22 10.08 MB/sec [4K blocks]
Uncached Read 536.72 99.59 MB/sec [256K blocks]

Ultimate 13.3 MBA RevC
Results 151.87
System Info
Xbench Version 1.3
System Version 10.6.4 (10F569)
Physical RAM 2048 MB
Model MacBookAir2,1
Drive Type APPLE SSD TS128B
CPU Test 113.33
GCD Loop 187.51 9.88 Mops/sec
Floating Point Basic 91.79 2.18 Gflop/sec
vecLib FFT 74.19 2.45 Gflop/sec
Floating Point Library 178.92 31.15 Mops/sec
Thread Test 209.61
Computation 213.57 4.33 Mops/sec, 4 threads
Lock Contention 205.78 8.85 Mlocks/sec, 4 threads
Memory Test 162.38
System 209.17
Allocate 271.23 996.04 Kalloc/sec
Fill 173.59 8440.14 MB/sec
Copy 204.29 4219.60 MB/sec
Stream 132.70
Copy 148.66 3070.43 MB/sec
Scale 150.52 3109.77 MB/sec
Add 159.44 3396.31 MB/sec
Triad 95.23 2037.18 MB/sec
Quartz Graphics Test 114.86
Line 101.80 6.78 Klines/sec [50% alpha]
Rectangle 125.94 37.60 Krects/sec [50% alpha]
Circle 107.55 8.77 Kcircles/sec [50% alpha]
Bezier 122.20 3.08 Kbeziers/sec [50% alpha]
Text 120.69 7.55 Kchars/sec
OpenGL Graphics Test 152.16
Spinning Squares 152.16 193.02 frames/sec
User Interface Test 113.61
Elements 113.61 521.41 refresh/sec
Disk Test 203.86
Sequential 132.54
Uncached Write 148.35 91.08 MB/sec [4K blocks]
Uncached Write 265.93 150.46 MB/sec [256K blocks]
Uncached Read 61.01 17.85 MB/sec [4K blocks]
Uncached Read 304.24 152.91 MB/sec [256K blocks]
Random 441.37
Uncached Write 257.93 27.30 MB/sec [4K blocks]
Uncached Write 358.33 114.72 MB/sec [256K blocks]
Uncached Read 1276.29 9.04 MB/sec [4K blocks]
Uncached Read 620.54 115.15 MB/sec [256K blocks]

 

[WardC]

I've taken the time to work with this data and compile these results in much more easy-to-read format, best result is shown in red:

 

[freitas]

Very nice summary

I've been using the 11 all day and one of the big things I have noticed, is how much cooler and quieter it runs compared to the 13.

 

[iRun26.2]

Very nice summary

I've been using the 11 all day and one of the big things I have noticed, is how much cooler and quieter it runs compared to the 13.

You have just emphasized a point that is very important to me:

The new 11.4" model sounds like it runs cooler and quieter than the new 13.3" model.

Many people may be drawn to the slightly faster performance of the 13.3" model. Me, I'd rather give up just a little bit of speed if my machine runs a little cooler (and, naturally, quieter too with less demand for the fan). A major advantage of the ULV part!

 

[Jobsian]

Terrific summaries WardC and freitas, many thanks. The rev c is performing better than expected. The 2.13 and extra 3mb l2 cache is worth quite a bit more than I thought. However as mentioned above, I completely agree, the temp profile is much better so far on the D.

 

[netdog]

You have just emphasized a point that is very important to me:

The new 11.4" model sounds like it runs cooler and quieter than the new 13.3" model.

Many people may be drawn to the slightly faster performance of the 13.3" model. Me, I'd rather give up just a little bit of speed if my machine runs a little cooler (and, naturally, quieter too with less demand for the fan). A major advantage of the ULV part!

The only time I've heard the fans on my new 2.13 was when I was running software update for some reason. Otherwise it's quite as can be for day to day tasks. I expect that for things like video compression, both models will rev up. but the 13 will leave the 11 in the dust. Of course, neither machine is ideal for rendering video.

 

[WardC]

Take into mind that the comparison benchmarks were comparing a 1.6GHz/4GB/128GB Flash 11.6" 2010 model to a 2009-model 2.13GHz/2GB/128GB SSD unit.

The 2009-model 2.13 smokes the new maxed out 11.6" in just about everything but the disk test, the new Flash NAND device is faster in most areas than the old SSD, but everything else that uses the processor is way faster on the 2.13