New MacBook Air Slower Than Older Model?

[benpatient]

MBA may be light, but it is too "big" in a length and width kind of way for me to ever consider one.

If it's going to be the size of a platter, it needs to have some ports on it.

I'm waiting for the 10" air.

 

[Hattig]

These Macworld tests are are a bunch of BS. See this thread, people consistently report higher Geekbench scores with MBA 2.13, as compared to older models:

https://forums.macrumors.com/threads/427141/

Interesting:

Geekbench Scores
3234 (MBA 2.13 Ghz running Snow Leopard WWDC build)
2682 (MBA 2.13 running 10.5.7)
2541 (MBA 2.13 running from restored Leopard image)
2485 (1.8GHz MBA w/ 1.8HDD)
2189 (MBA 1.6 80GB Rev. A)

 

[Victor Odin]

And yet it benchmarks slower. Do you see how that's a) inexplicable, and b) not at all related to any point you made?

No.

Heat is almost always the limiting factor in laptop design.

This phenomenon can be 100% explained by thermal management (or a lack thereof).

Unless you are 100% positive that nothing has changed with regards to thermal management between generations, then you cannot draw a linear equation of faster rated processor = faster computer.

CPU's don't always run at their rated speed.

CPU's run slower as they heat, regardless of clock speed.

There are also issues of electromigration with heat, but until you can accept the premise that Apple doesn't always do the perfect thing, the explanation will be lost on you.

 

[polaris20]

The Air is a second computer.

Highly portable for web browsing and MS Word types of usage. It's the closest Apple will get to the netbook form. I don't think anyone should expect to do any photo/video/audio editing on it!

I would argue that you can still do audio and light video on it too. I've been editing video and audio all the way back to the PIII 1Ghz days.

 

[bananas]

I'm getting another MBA, this time with SSD. This thing is just great. It's much more powerful than Lenovo X301, which is the second most interesting laptop IMHO. The only thing that is better in the X301 is the resolution. I hope that Apple will soon upgrade the MBA to 1440x900 too.

EDIT: and while you're upgrading it, please make it 4 GB RAM.

 

[rumody]

This is how it should look like

Form over function laptop.

Totally agree. I think they should keep the brick-form factor of the pro unibodies, and just try to make it thinner. Without a CD, that will be a great portable computer:

http://macbookair.rumody.com/

 

[derek1984]

Why I never considered buying a MBA: No optical drive. And the newer Macbooks are thin and light enough.

 

[andreab35]

Haha, I just looked at this and laughed.
I don't know how this can be true.

You think performance will change for these laptops once Snow Leopard rolls out?

 

[bananas]

so the bad benchmark result must be due to SATA firmware issues that MacBook Pro had too? I'm sure it will be fixed soon.

 

[*LTD*]

Yeah I'm sure an Adamo would be a way better choice.

Why? It's a Windows box.

 

[applecultvictim]

Macworld are slightly less bozos than famous bozos anandtech.

They must have messed up here.

Or intel hasn't got the thermals right and apple had to cut down to allow it to perform well, what with the gen A fiasco not wanting to repeat.

In either case it's either a fluke result or something apple had to do.

And more so these tests don't come close to real life use.

Also it's kinda stupid to call an ultraportable underpowered, which ultra portable is NOT underpowered you bozos? It comes with the territory.


The air is still the best computer for all the on the go professionals or students, stylish, fast (relative to the competition) with a marvelous screen and now it's also a very very good deall pricewise. Nothing in the pc world can even come close.

It's a winner in all fronts, and it's not apple's fault that quadcores have not become availble for ultraportables.

 

[50548]

How can they be slower

Unclear to me as well; it's easier to say that Speedmark scores are extremely fluid and NEVER to be trusted.

 

[Shadow]

Every computer is different. They all have different thermal profiles.

You friend's MBA may run faster or slower than yours. Every piece of silicon is different.

If one runs hotter than the other, then it may be slower.

Here's some more smilies for you

Actually I would argue exactly the opposite. Every piece of silicon is pretty much identical to similar units; otherwise the failure rate would be astronomical. Granted, there are small variations between machines, but this is to do with the environment they're in and the software they're running rather than significant differences between similar chips and whatnot.

 

[50548]

so the bad benchmark result must be due to SATA firmware issues that MacBook Pro had too? I'm sure it will be fixed soon.

Since the very beginning of that story it was said that the new MBA was NOT affected by these issues...

 

[Victor Odin]

It's a winner in all fronts, and it's not apple's fault that quadcores have not become availble for ultraportables.

Apple doesn't have quadcores for anything portable at all.

Not even the 17" Mac Book Pro.

Heck, not even iMac, and that's not even portable.

It's not Apple's fault, though.

It's the mean magic wizard's fault. It's a never ending struggle between good and evil, and Apple is tirelessly fighting to absolve the world of sin and wrongdoing.

Kind of like Jesus, only way better.

 

[applecultvictim]

Apple doesn't have quadcores for anything portable at all.

Not even the 17" Mac Book Pro.

Heck, not even iMac, and that's not even portable.

It's not Apple's fault, though.

It's the mean magic wizard's fault. It's a never ending struggle between good and evil, and Apple is tirelessly fighting to absolve the world of sin and wrongdoing.

Kind of like Jesus, only way better.

Thanks for the irony, I kinda liked it despite it mocking me.

On a serious note however I said ULTRAPORTABLE not portable, and quads are not available there. Apple has put THE BEST ultraportable cpu they could get their hands on in the air. Is there anything else they could have done, tell me?

Exactly. So how is it their fault?

 

[hitekalex]

Thanks for this. I was in the market for the 2.13 but was very much concerned about the Macworld tests. I was esp. pleased to see the score of the MBA under Snow Leopard...

Yep, typing this on MBA 2.13 running Snow Leopard. The thing flies.

 

[Gasu E.]

Please! This is Apple! It's not "slow", it's "craftsmanlike."

 

[Victor Odin]

Actually I would argue exactly the opposite. Every piece of silicon is pretty much identical to similar units

And you would be wrong. Most CPU's are stamped from the same sheet yet are highly variable. The ones that can tolerate more heat are rated for higher speeds. The ones that can tolerate less heat are rated for slower speeds. There are a large amount that are simply discarded.

otherwise the failure rate would be astronomical.

It is fairly high. Not all CPU's are put into machines. They are tested, and if they fail, they test them at a slower speed. If they pass at that speed, that is the speed at which they are rated.

All processors are more or less the same die and are theoretically capable of running at the highest clock speed, but are extremely variable.

The ones that "fail" are slowed down until they no longer fail and sold to OEM's or equipment makers.

This happens even when they come out of the exact same fab.

Granted, there are small variations between machines, but this is to do with the environment they're in and the software they're running rather than significant differences between similar chips and whatnot.

You are incorrect.

It is also why true benchmarking is hard unless you use a large sample.

The same CPU can run 5-10C hotter than another, and still be within themal spec. The hotter one will throttle more quickly though and will be more prone to electromigration.

Fortunately, most people will not notice.

Will your new MBA be faster than the old one?

Maybe. It's entirely possible.

Then agin, maybe not.

Unless Apple is garantee that you will see x% better performance in a particular app, you won't be able to do anything anyway.

The point is that I don't think the people who did the tests are rabid Mac bashers. I don't think this was some large conspiracy to crap on the MBA.

To the contrary, they like it.

Perhaps they also had a hotter than usual chip. They should have posted temps. This would have helped us figure it out.

The outcome of the test is also EASILY explainable for those that care to consider the explanation.

Some will hear nothing of it, though.

There is logic, and there is religion.

Mac users are kind of split down the middle as to how they approach things.

 

[Victor Odin]

Thanks for the irony, I kinda liked it despite it mocking me.

On a serious note however I said ULTRAPORTABLE not portable, and quads are not available there. Apple has put THE BEST ultraportable cpu they could get their hands on in the air. Is there anything else they could have done, tell me?

Exactly. So how is it their fault?

The point was ... even if there was an ultra-portable quadcore, I have no idea why you think they would use it when such chips are available for other laptops ... yet are missing from all portable Apple products.

If Apple had any interest whatsoever in portable quad-core, they could have accomplished it.

Obviously, they do not.

I don't know that it's a "fault", but it is a choice that they make.

 

[applecultvictim]

The point was ... even if there was an ultra-portable quadcore, I have no idea why you think they would use it when such chips are available for other laptops ... yet are missing from all Apple products.

Yeah but that's hypothetical, you can't hold them responsible with the logic that they are doing so and so, but had the situation been different (available ultraportable quads) they would be doing this and that, and this and that would be wrong. You just can't judge someone like that.

But even going along with*your logic. Both the air and the iphone are very sharply placed in terms of tec, the imac isn't. Thus the former would get all the latest and best hardware while the imac would again be as is more slower in adoption. But to tell you the truth for most people a quad vs. a dual core imac wont make much difference while it makes a ton of difference in ultra mobile apps, hence apple are quick there.

 

[Stephen123]

It's both funny and sad how folks are arguing about the MacWorld tests being valid or not, and the strange relationship between clock speed and benchmarks while completely ignoring the Geekbench results being posted here which contradict the MacWorld results.

 

[nick9191]

Smells like bull to me.

The old Rev B used the SL9300 in the low end, and the SL9400 in the high end. The new model uses the SL9400 in the low end, and the SL9600 in the high end.

How can it possibly be slower? Same chip generation, presumably same board.

 

[Victor Odin]

Yeah but that's hypothetical, you can't hold them responsible with the logic that they are doing so and so, but had the situation been different (available ultraportable quads) they would be doing this and that, and this and that would be wrong. You just can't judge someone like that.

Are you an imbecile or do you just play one on the Internet?

There is no evidence for your premise. None. There is a great deal of evidence for mine, however.

It's not hypothetical, Apple does not have quad-cores in their highest end $4,000 laptops. There is no quad-core option. This is not theory. This is fact.

Can we "blame" Apple for that?

The fastest, top-of-the-line Mac "pro" portable has no quad CPU, when quads are obviously available, and have been for some time. Even the newest 17" portable has no quad.

This being the case, your assertion that they would put quads in the Mac Book Air is patently absurd on it's face.

But even going along with*your logic. Both the air and the iphone are very sharply placed in terms of tec, the imac isn't.

The iPhone is just now getting features that I have been enjoying for 4 years on other devices.

Ergo, I disagree with your premise.

I'm still enjoying the hell out of my A2DP, as I have been since 2006.

Are you?

Wait ... Apple hasn't told you that you want it yet.

You'll probably like it when they "invent" it, though.

Thus the former would get all the latest and best hardware while the imac would again be as is more slower in adoption. But to tell you the truth for most people a quad vs. a dual core imac wont make much difference while it makes a ton of difference in ultra mobile apps, hence apple are quick there.

Then why doesn't one, single, solitary Mac laptop have one?

 

[50548]

Smells like bull to me.

The old Rev B used the SL9300 in the low end, and the SL9400 in the high end. The new model uses the SL9400 in the low end, and the SL9600 in the high end.

How can it possibly be slower? Same chip generation, presumably same board.

As I said above, Speedbench is NOT reliable, as its final results do NOT reflect real-life usage and are widely subject to variations...kinda like XBench actually.