MacBook Air 1.6Ghz HDD vs 1.8Ghz SSD Benchmarks

[Maxintech]

100 Mb/s would be quite pathetic. Fortunately, Ultra DMA mode 5 runs at 100 MB/s, that is eight times as fast as you say. (Mb = Megabit, MB = Megabyte).

I known the difference. It's was a typewrite mistake. Thanks for the correction.

 

[ncbill]

Ouch!

That means in 3 years, we're still paying the same $999 for the SSD upgrade, though we'll get 128GB instead of 64GB.

And those new, fast Samsung SSDs have the SATA interface - sure we'll be able to drop them in the current MBA?

My guess would be $10/GB for large capacity SSDs in 3 years or so.

 

[AidenShaw]

...but this is true for any laptop, and even if you had a spare battery then what do you do after 6 hrs prey tell me?

On my 8 hour flight, I'd simply swap in the second spare battery that I brought along.

And, by the way, my solution is scalable. If I had a 10 hour flight (or conference), I'd bring a third spare. It isn't hard to figure out how many spare batteries that you need to go "n" hours without a power strip.

Unless I had a MacBook Cube Air - then I'd be [expletive not typed, insert you own crude gerund].

I would use my iPhone for movie watch on the long flight.

Movies? Typically on the flight to the meeting I'm merging two to five Powerpoint presentations and editing to get the right message for the customer that I'll be meeting at 08:00 in the morning.

Watch a movie? LOL, I wish that I had the free time.

...and at conferences there's always a power strip there's always a power strip so no worry there either. switched out for another. This is simply a non-issue for me.

In small meetings, power usually isn't a problem. Put a couple of thousand people at a keynote, though - I've never seen a power strip at every chair in the auditorium. Same for the conferences where you have a couple of hundred people at tables.

Which conferences have you been at where "there's always a power strip"? We obviously have not met.

 

[MathiasMag]

What about parallel I/O

All of this seems to be mainly concerned with single stream processing. In the real world, you're likely to not use one app at a time.

How about playing music from MP3s on the disk, while email is downloading and a very large picture is being saved.

I'd like to see a comparison of say five parallel jobs doing parallel I/O. Seeing each category (reads/writes/sequential/random) and then one where it is a complete mix of all would be very interesting. This is probably where the SSD would show it's true value.

Another interesting test would be to have a lot going on and the machine being out of memory so the swap area gets used a lot. If the memory pressure is enough, a disk based system would be thrashing. I wonder if the SSD would survive better and as a result you could have less memory and still have acceptable performance.

The real interesting thing here is not what we pay today, it is the fact that this is where disk technology is going and HD will not be part of computing for that much longer. No introduction of new technology has ever been done at budget prices. However, the cost of SSD will eventually be lower than HD and the speed higher. This is clearly the future of high speed I/O, especially when you have multiple accesses in parallel for the same device.

 

[Rocketman]

Apple, it seems, is using one of the slowest SSD's out there. There are SSD's with transfer rates that could saturate the PATA bus available, or soon to be. That's where we're headed, and that's exciting. Standard hard drives are only going to get faster as the density of the platters increases, and the prospects of that aren't all that great, are they?

Apple for the longest time stayed at USB 1.1 to give 1394-A Firewire a better adoption rate among Mac users. Many of those legacy system owners now sure wish Apple went to USB 2.0 much sooner, simply because they could have.

I suspect the slower SSD is a supplier issue. Existing contracts for components, rather than go to the newest stuff and pay full price. Apple has a new profit margin emphasis (very good for stock and future business prospects). Steve was somewhat dragged kicking and screaming into releasing a product with SSD. To me this indicates he has a longer term plan that involves a device with a cost/performance/package sweet spot.

What I find amusing is still to this day, the OS and applications I ran on my old Macintosh Plus in ram disc (saving to a SCSI drive externally) were practically much faster than the bloatware running on todays much faster computers. Notably modern stuff we do also pushes millions more pixels per minute too.

Rocketman

 

[Manic Mouse]

2.2ghz MacBook w/ Hitachi 200GB vs. 1.8ghz MacBook Air w/ 64GB SSD

I'll let the numbers speak for themselves. MBA figures in parentheses.

CPU Test 113.26 (99.61)
Thread Test 185.00 (134.99)
Memory Test 157.54 (148.00)
Quartz Graphics Test 171.20 (107.74)
User Interface Test 235.54 (113.53)
Disk Test 41.40 (47.26)

Sequential 71.63 (40.82)

Uncached Write 107.36 65.92 MB/sec [4K blocks] (20.83MB/s)
Uncached Write 105.57 59.73 MB/sec [256K blocks] (26.32MB/s)
Uncached Read 34.26 10.03 MB/sec [4K blocks] (7.97MB/s)
Uncached Read 127.12 63.89 MB/sec [256K blocks] (48.75MB/s)​

Random 29.11 (56.13)

Uncached Write 9.57 1.01 MB/sec [4K blocks] (2.23MB/s)
Uncached Write 90.98 29.13 MB/sec [256K blocks] (16.92MB/s)
Uncached Read 72.73 0.52 MB/sec [4K blocks] (7.02MB/s)
Uncached Read 123.05 22.83 MB/sec [256K blocks] (48.24MB/s) ​

Price as tested: $1446.00 ($2868.00) <-- Academic pricing, both systems. Did the Air a favor and didn't include tax.

Results	109.44	
	System Info		
		Xbench Version		1.3
		System Version		10.5.1 (9B18)
		Physical RAM		4096 MB
		Model		MacBook3,1
		Drive Type		Hitachi HTS722020K9SA00
	CPU Test	113.26	
		GCD Loop	259.62	13.69 Mops/sec
		Floating Point Basic	123.36	2.93 Gflop/sec
		vecLib FFT	73.91	2.44 Gflop/sec
		Floating Point Library	101.76	17.72 Mops/sec
	Thread Test	185.00	
		Computation	197.26	4.00 Mops/sec, 4 threads
		Lock Contention	174.18	7.49 Mlocks/sec, 4 threads
	Memory Test	157.54	
		System	163.50	
			Allocate	249.85	917.55 Kalloc/sec
			Fill	134.78	6553.18 MB/sec
			Copy	144.36	2981.74 MB/sec
		Stream	152.00	
			Copy	142.54	2944.15 MB/sec
			Scale	142.73	2948.80 MB/sec
			Add	163.20	3476.48 MB/sec
			Triad	162.18	3469.47 MB/sec
	Quartz Graphics Test	171.20	
		Line	161.96	10.78 Klines/sec [50% alpha]
		Rectangle	210.23	62.76 Krects/sec [50% alpha]
		Circle	166.54	13.57 Kcircles/sec [50% alpha]
		Bezier	165.61	4.18 Kbeziers/sec [50% alpha]
		Text	160.48	10.04 Kchars/sec
	User Interface Test	235.54	
		Elements	235.54	1.08 Krefresh/sec
	Disk Test	41.40	
		Sequential	71.63	
			Uncached Write	107.36	65.92 MB/sec [4K blocks]
			Uncached Write	105.57	59.73 MB/sec [256K blocks]
			Uncached Read	34.26	10.03 MB/sec [4K blocks]
			Uncached Read	127.12	63.89 MB/sec [256K blocks]
		Random	29.11	
			Uncached Write	9.57	1.01 MB/sec [4K blocks]
			Uncached Write	90.98	29.13 MB/sec [256K blocks]
			Uncached Read	72.73	0.52 MB/sec [4K blocks]
			Uncached Read	123.05	22.83 MB/sec [256K blocks]

Ouch, the MBA got it's ass kicked by a computer that cost half as much. But hey, it's .5" thinner and has the "feature" of no disc drive!

So if it doesn't make a good portable/business computer (so-so battery life, non-replaceable battery), is really expensive and isn't all that powerful just what is the point of the MBA? Is it just for people with more money than sense?

 

[uNext]

Ouch, the MBA got it's ass kicked by a computer that cost half as much. But hey, it's .5" thinner and has the "feature" of no disc drive!

So if it doesn't make a good portable/business computer (so-so battery life, non-replaceable battery), is really expensive and isn't all that powerful just what is the point of the MBA? Is it just for people with more money than sense?

I think it is aimed at people with more money then sense.
There is just no point to it. Benchmarks all over the internet
are proving what we all know that the computer just sucks. But what is amazing is how people try to justify the purchase by saying "if you dont like it it was not meant to be for you" but some of those people are quick to bash windows vista wouldn't the same rule apply?

The macbook topped it and is way cheaper with more features. The MBair is the razr of laptops. style over function.

 

[50548]

And you never are away from an electrical outlet for more than 3 hours (or you don't care if the battery goes dead on you until you can find an outlet)....

A real weak point, though, is the lack of a user-swappable battery so that one could carry a spare battery or two for the long flight or all-day meeting at a conference.

It's a shame when your laptop dies before lunch, and you can't simply swap a spare battery in and keep on going.

Aiden, two simple facts for pretty much EVERY professional meeting out there:

1 - People use notebooks CONNECTED TO power outlets in meetings, not the opposite;

2 - There is NO need at all for Ethernet or any other physical connection; it's all about Wi-Fi.

You probably don't attend that many meetings, otherwise you would know that. Battery time is useful for airplanes or in transit...nothing else.

And no, nobody carries spare batteries, this has been a bogus argument from the outset.

 

[uNext]

Aiden, two simple facts for pretty much EVERY professional meeting out there:

1 - People use notebooks CONNECTED TO power outlets in meetings, not the opposite;

2 - There is NO need at all for Ethernet or any other physical connection; it's all about Wi-Fi.

You probably don't attend that many meetings, otherwise you would know that. Battery time is useful for airplanes or in transit...nothing else.

And no, nobody carries spare batteries, this has been a bogus argument from the outset.

You feel comfortable with having a wifi connection during an important presentation?

 

[DeathChill]

I'm sure it's already been said but XBench is extremely unreliable. You can get vastly different scores on the same machine with no changes made to the test circumstances.

 

[arn]

Ouch, the MBA got it's ass kicked by a computer that cost half as much. But hey, it's .5" thinner and has the "feature" of no disc drive!

I don't come to that conclusion at all. Of course CPU speed of a 2.2GHz vs a 1.8GHz is going to be faster in the MacBook, but the key benchmark here is of disk access, and the MBA was faster by a large margin for the key metric that it's supposed to be faster in. non-sequential reads.

Uncached Write 1.01 MB/sec [4K blocks] (2.23MB/s)
Uncached Write 29.13 MB/sec [256K blocks] (16.92MB/s)
Uncached Read 0.52 MB/sec [4K blocks] (7.02MB/s)
Uncached Read 22.83 MB/sec [256K blocks] (48.24MB/s)

Bold is faster. Parens are MBA.

So for random reads, the MBA is anywhere from 2x to 13.5x faster. It is also faster for random writes for small 4k blocks.

arn

 

[ViveLeLivre]

Ouch, the MBA got it's ass kicked by a computer that cost half as much. But hey, it's .5" thinner and has the "feature" of no disc drive!

So if it doesn't make a good portable/business computer (so-so battery life, non-replaceable battery), is really expensive and isn't all that powerful just what is the point of the MBA? Is it just for people with more money than sense?

Of course, the greatest irony is that the MBA actually has a larger footprint than the MacBook. Also, MBA advocates conveniently forget that hard drives in the MacBook can be upgraded very easily -- there is nothing to prevent MB (or MBP) owners from upgrading to 2.5" SSD drives when they become available.

I honestly thought the MBA's SSD performance would trounce anything the MacBook could throw at it. After seeing these numbers I am very pleased with my purchase.

 

[ViveLeLivre]

I don't come to that conclusion at all. Of course CPU speed of a 2.2GHz vs a 1.8GHz is going to be faster in the MacBook, but the key benchmark here is of disk access, and the MBA was faster by a large margin for the key metric that it's supposed to be faster in. non-sequential reads.

Uncached Write 1.01 MB/sec [4K blocks] (2.23MB/s)
Uncached Write 29.13 MB/sec [256K blocks] (16.92MB/s)
Uncached Read 0.52 MB/sec [4K blocks] (7.02MB/s)
Uncached Read 22.83 MB/sec [256K blocks] (48.24MB/s)

Bold is faster. Parens are MBA.

So for random reads, the MBA is anywhere from 2x to 13.5x faster. It is also faster for random writes for small 4k blocks.

arn

Sure, for RANDOM reads... and the Hitachi is up to 3x as fast for everything else (except 4k random writes).

For $2900+, it ought to wipe the floor with my MacBook in every category.

 

[OS X Dude]

I'm sure it's already been said but XBench is extremely unreliable. You can get vastly different scores on the same machine with no changes made to the test circumstances.

I hear that - under the same circumstances my MacBook scored 3 points lower than it did on the first test :S

I suppose it's there for a ballpark figure only - do like 5 or more tests and take an average from those results.

"Of course, the greatest irony is that the MBA actually has a larger footprint than the MacBook. Also, MBA advocates conveniently forget that hard drives in the MacBook can be upgraded very easily -- there is nothing to prevent MB (or MBP) owners from upgrading to 2.5" SSD drives when they become available."

Nice - now I dunno whether to plump for a 7200rpm HDD now or wait 4/5 years till my next mac haha.

 

[JGowan]

Question is... "Would you really like to pay $1,000 for it?"

I wonder how many would.

Cheers!

$1000 is relative. To a beggar, it's a fortune. To a businessman, it's a paltry sum. (Consider M$ bid of $44.6B --> they think Yahoo! is worth it. Amazing.)

…

Some people look at computer purchases differently than others. Many buying a computer know that they will be buying a new one every 2-3 years so they don't bother with the absolute best. Others would like to make their purchase last for 5-6 years so they buy the top of the line. I tend to be in the later category. My desktop (Feb '02) and laptop (Feb '03) are due for an upgrade. I plan on dropping about $15K on hardware/software at that time so another thousand would not be a big deal. I make my living on this stuff so I want (and feel I deserve) the best. That being said, my next laptop will be a 15" MBP not a MBA.

Also,… for me, I would have a hard time knowing that the hard drive I use for my iPod Classic is the same one in the laptop I plan on getting. It just seems too slow to run a machine a lot bigger and more complicated than an iPod.

 

[JonnyMac]

I think it is aimed at people with more money then sense.
There is just no point to it. Benchmarks all over the internet
are proving what we all know that the computer just sucks. But what is amazing is how people try to justify the purchase by saying "if you dont like it it was not meant to be for you" but some of those people are quick to bash windows vista wouldn't the same rule apply?

The macbook topped it and is way cheaper with more features. The MBair is the razr of laptops. style over function.

I'm not one of the MBA advocates saying "if you don't like it it was not meant to be for you." I don't like Vista and I prefer OS X over it but I don't bash Vista, I simply don't buy it. I migrated from a MacBook and I am enjoying the smaller size of the MBA. Granted, I am not a power user by any means and I only use it to watch movies, surf the Internet, email, travel with, etc. It all comes down to if you're happy with your purchase and if it's worth it to YOU, the owner. SSD isn't worth it to me, but it could very well be for other users. If the MBA is not worth it to you, don't buy it. Buy a MB or MBP or a PC laptop, if those computers match your needs more...

 

[arn]

For $2900+, it ought to wipe the floor with my MacBook in every category.

Only if you value processing power over all else. And place no value in added portability.

I know this is a tired argument, but one last time.

1. 8Core Mac Pro. $2799.
2. 17" MacBook Pro. $2799.

The Mac Pro beats the MacBook Pro across the board on benchmarks. Therefore, by your logic, no one should buy a MacBook Pro. And if you are going to attach the computer to your desk, then the MacPro is more bang-for-your-buck. If you ever want to leave your desk with your computer. You'll gladly pay the $XXX premium for a MacBook Pro rather than try to lug around the Mac Pro where you want.

Similarly, for some people the extra 2 pounds of savings is worth the $XXX-$XXXX premium of the MacBook Air. Especially if they don't use their computer for heavy duty processing applications.

So while a 2 pound difference makes no difference to you if all you are doing is carrying around you MacBook from home to your parents christmas, if you are carrying it around every day with you, the 2 pounds is a big enough difference that it's worth $XXX.

arn

 

[Silentwave]

I'm sure it's already been said but XBench is extremely unreliable. You can get vastly different scores on the same machine with no changes made to the test circumstances.

I agree. At least on Leopard, I'm almost never able to get the same score more than once. I've had variabilities of over 20 points on the overall, and several dozen points on individual tests. I think the worst was when I had something similar to the MBAir SSD tested earlier in the thread, where the UI test managed to *double* with the next run.

Then, when comparing with a friend who has the previous generation 2.16 C2D MBP and 3GB RAM, plus a slower hard drive and older GPU, running Tiger, my current generation MBP 2.4/4GB system with a 7200RPM HDD, the new faster GPU with double the VRAM running Leopard never once beat his XBench score...only tied it once. If the benchmark were accurate, I'd be seeing at least some benefit from the faster FSB, additional RAM, and better graphics, and the slightly higher clock speed (it wouldn't be much, but it would be something).

XBench needs an update badly. 1.3 came out back in August 2006. It's now February 2008. In that time we've gone to 8 cores, several new GPUs, and needless to say a new OS. Other apps such as Geekbench have updated numerous times and are more consistent.

 

[arn]

https://forums.macrumors.com/posts/4881470/

posted by ciparis

More tests, MBP vs MBA this time:

Launch 4 apps after booting (iPhoto, iMovie, iDVD, GarageBand).

MacBook Pro (about 1:30)
http://qik.com/video/14849

MacBook Air (11 seconds):
http://qik.com/video/14862

 

[ViveLeLivre]

Only if you value processing power over all else. And place no value in added portability.

I know this is a tired argument, but one last time.

1. 8Core Mac Pro. $2799.
2. 17" MacBook Pro. $2799.

The Mac Pro beats the MacBook Pro across the board on benchmarks. Therefore, by your logic, no one should buy a MacBook Pro. And if you are going to attach the computer to your desk, then the MacPro is more bang-for-your-buck. If you ever want to leave your desk with your computer. You'll gladly pay the $XXX premium for a MacBook Pro rather than try to lug around the Mac Pro where you want.

Similarly, for some people the extra 2 pounds of savings is worth the $XXX-$XXXX premium of the MacBook Air. Especially if they don't use their computer for heavy duty processing applications.

So while a 2 pound difference makes no difference to you if all you are doing is carrying around you MacBook from home to your parents christmas, if you are carrying it around every day with you, the 2 pounds is a big enough difference that it's worth $XXX.

arn

You're right, that strawman is tired.

You can't compare the .5" difference in width between the MB and MBA with the difference in volume and weight between a desktop tower and a laptop. The MB and MBA literally compete in the same space. The MBP and Mac Pro do not.

If width and weight are all that matters to you, fine. But by any other objective measure, the MacBook still kicks the crap out of the MBA.

 

[arn]

You can't compare the .5" difference in width between the MB and MBA with the difference in volume and weight between a desktop tower and a laptop. The MB and MBA literally compete in the same space. The MBP and Mac Pro do not.

all I can do is quote this

http://www.macworld.com/article/131864/2008/01/macbookair.html

One reason I loved the 12-inch PowerBook G4 was that it crossed some hard-to-define weight barrier, one I hadn’t even been aware of until I started using a laptop that crossed it. The 12-inch PowerBook was so small and light that carrying my laptop around with me became an afterthought. Instead of lugging a 15-inch PowerBook from place to place, I could idly hold the 12-inch model in one hand. The MacBook Air takes that easy feeling to an extreme

 

[ViveLeLivre]

https://forums.macrumors.com/posts/4881470/

posted by ciparis

Just ran this test on my MacBook on 5% battery power: took 9.2 seconds.

 

[ViveLeLivre]

One reason I loved the 12-inch PowerBook G4 was that it crossed some hard-to-define weight barrier, one I hadn’t even been aware of until I started using a laptop that crossed it. The 12-inch PowerBook was so small and light that carrying my laptop around with me became an afterthought. Instead of lugging a 15-inch PowerBook from place to place, I could idly hold the 12-inch model in one hand. The MacBook Air takes that easy feeling to an extreme

It was also 12", which is what the MBA should have been. Apple could have stuck a Celeron in a 12" laptop and I wouldn't be criticizing it's performance. Instead they had to make the thinnest 13.3" notebook evah! with little thought to or total disregard for the actual functionality and performance of such of a device. 2lbs. and half an inch doesn't save it when it fits in the exact same slot in my backpack! I might as well spend half as much, get 1.5x the same computer and learn to deal with the extra weight.

A 12" would have been light, useable, insanely portable, and still very thin if they can shove even that gimpy computer into the MBA's chassis. Instead we get a metal MacBook wannabe with a crew cut, no optical disk -- and light-up keys. Brilliant.

 

[elmo151]

If width and weight are all that matters to you, fine. But by any other objective measure, the MacBook still kicks the crap out of the MBA.

there is no reason to buy an MBA if you don't want one. so why spend so much time bitching about it.

It fits some peoples needs(mine) and I bought one. If it doesn't fit yours; don't buy it..

 

[Gunga Din]

there is no reason to buy an MBA if you don't want one. so why spend so much time bitching about it.

It fits some peoples needs(mine) and I bought one. If it doesn't fit yours; don't buy it..

i think most people are just upset because this product took time and development away from the Macbook Pro Update.

I'm looking into picking up my first Mac ever and during that keynote speech when Jobs intrduced the MacAir, I honestly thought it was some kind of joke lol. I was looking behind him for the real computer he was gonna introduce. This was the first time i've watched a keynote so i wasnt sure how he did things.

But anyway, yea when he introduced the MacAir, i swear i heard crickets in the crowd and i said to myself.....ehh?

Maybe if I owned a MacPro or something already , I could somehow consider it but the SSD thing is too new for release and incorporation into a notebook as far as i'm concerned. Maybe a yr or 2 from now, but this timing seemed off to me. 3k? No.