PDA

View Full Version : 1.6 Rev B faster than 1.6 Rev A?




Kan-O-Z
Dec 3, 2008, 03:09 PM
I think it will be overall faster(bus speed, ram speed, better video) but my question is this:

Is the CPU faster than the old one? I know that they are using a Penryn 45nm version now. Will the new 1.6 be comparable to the older 1.8. Again I am just referring to CPU.

The reason I ask this is that Penryn saw a 5-10% boost over Merom for the exact same clock speed I believe.

Kan-O-Z



nick9191
Dec 3, 2008, 03:18 PM
Yes. The cache is double on the new model (6mb vs. 3mb).

Ironically it has a larger cache than all 3 Macbooks, and even the baseline pro.

Kan-O-Z
Dec 3, 2008, 03:55 PM
Yes. The cache is double on the new model (6mb vs. 3mb).

Ironically it has a larger cache than all 3 Macbooks, and even the baseline pro.

But does anyone know if the processor itself is faster (not because of the cache, not because of the bus speed or RAM speed). Is the Penryn 1.6GHz faster than the 1.6GHz older processor.

Now that I think about it, it would be nearly impossible to know this as you can't test a Penryn with the old cache, bus speed and ram ;)

Kan-O-Z

Pixellated
Dec 3, 2008, 03:57 PM
But does anyone know if the processor itself is faster (not because of the cache, not because of the bus speed or RAM speed). Is the Penryn 1.6GHz faster than the 1.6GHz older processor.

Now that I think about it, it would be nearly impossible to know this as you can't test a Penryn with the old cache, bus speed and ram ;)

Kan-O-Z

No. Apart from the cache, DDR3, FSB etc...

Molopo
Dec 3, 2008, 04:51 PM
I think common sense would dictate that any two processors with identical core speeds would the same speed. Your newfound interest in macbook airs is good and all, but try to keep your questions reasonable...

Mactagonist
Dec 3, 2008, 05:04 PM
But does anyone know if the processor itself is faster (not because of the cache, not because of the bus speed or RAM speed). Is the Penryn 1.6GHz faster than the 1.6GHz older processor.

Now that I think about it, it would be nearly impossible to know this as you can't test a Penryn with the old cache, bus speed and ram ;)

Kan-O-Z

It is faster because of all those things. Otherwise they are the same architecture.

andyOSX
Dec 7, 2008, 08:42 AM
Uhhhh, no you guys.

The new MBAs use a 45nm penryn vs a shrunken down 65nm merom in the Rev As. A 45nm processor will always perform better at the same clock speed while simultaneously producing less heat. So yes, JUST the processor in a new MBA is faster at the same clock b/c it is a 45nm process.

nick9191
Dec 7, 2008, 09:17 AM
But does anyone know if the processor itself is faster (not because of the cache, not because of the bus speed or RAM speed). Is the Penryn 1.6GHz faster than the 1.6GHz older processor.

Now that I think about it, it would be nearly impossible to know this as you can't test a Penryn with the old cache, bus speed and ram ;)

Kan-O-Z
The cache is part of the CPU.

The CPU is faster because it has a larger cache.

And because of the more efficient 45nm architecture.

Kan-O-Z
Dec 7, 2008, 10:00 AM
I wonder if the new 1.6 CPU is as 'fast' as the old 1.8 CPU when just comparing CPUs and nothing else?

Kan-O-Z

r6girl
Dec 7, 2008, 10:09 AM
Uhhhh, no you guys.

The new MBAs use a 45nm penryn vs a shrunken down 65nm merom in the Rev As. A 45nm processor will always perform better at the same clock speed while simultaneously producing less heat. So yes, JUST the processor in a new MBA is faster at the same clock b/c it is a 45nm process.

That doesn't make any sense, unless the processor industry applies the same logic to defining a processor's clock speed as the hard drive industry does to specifying the capacity of an empty hard drive.

If one processor is faster than another, it should have a different clock speed, regardless of the architecture.

A Corvette traveling at 55mph is not going any faster than a Beetle at 55mph just because it has a larger and more powerful engine.

Kan-O-Z
Dec 7, 2008, 10:54 AM
That doesn't make any sense, unless the processor industry applies the same logic to defining a processor's clock speed as the hard drive industry does to specifying the capacity of an empty hard drive.

If one processor is faster than another, it should have a different clock speed, regardless of the architecture.

A Corvette traveling at 55mph is not going any faster than a Beetle at 55mph just because it has a larger and more powerful engine.

I disagree with you. Think of clock speed as rpm for a car engine. The RPMs of a car do no determine it's horsepower. The same with CPUs.

When new CPUs come along there are architecture changes. Sometimes maybe slight changes like between current Penryn and previous Merom. Sometimes bigger changes like Core 2 Duo v. Core Duo. The latest architecture will often be more efficient per clock cycle so in essence it is faster at the same clock speed.

As an example, a current 2.0GHz core 2 duo is probably the equivalent of a 5.0GHz Pentium 4 ;)

Kan-O-Z

r6girl
Dec 7, 2008, 11:01 AM
I disagree with you. Think of clock speed as rpm for a car engine. The RPMs of a car do no determine it's horsepower. The same with CPUs.

When new CPUs come along there are architecture changes. Sometimes maybe slight changes like between current Penryn and previous Merom. Sometimes bigger changes like Core 2 Duo v. Core Duo. The latest architecture will often be more efficient per clock cycle so in essence it is faster at the same clock speed.

As an example, a current 2.0GHz core 2 duo is probably the equivalent of a 5.0GHz Pentium 4 ;)

Kan-O-Z

I understand what you mean, but you asked if there is a difference in speed in processors technically rated as having the same clock speed. It's clear that they'll operate differently because of their different architecture, but if they're both 1.6Ghz processors, they'll both technically run at 1.6Ghz. Again, Corvette at 55mph vs. a Beetle at 55mph - same speed, though the Corvette is obviously more powerful and capable of much more. Your question was whether or not the Corvette is faster at the same speed - you didn't ask about efficiency or capacity, which are different than speed.

Probably just semantics at this point - I do that sometimes... :p

nick9191
Dec 7, 2008, 11:02 AM
That doesn't make any sense, unless the processor industry applies the same logic to defining a processor's clock speed as the hard drive industry does to specifying the capacity of an empty hard drive.

If one processor is faster than another, it should have a different clock speed, regardless of the architecture.

A Corvette traveling at 55mph is not going any faster than a Beetle at 55mph just because it has a larger and more powerful engine.

http://uk.youtube.com/watch?v=PKF9GOE2q38

Watch

r6girl
Dec 7, 2008, 11:31 AM
http://uk.youtube.com/watch?v=PKF9GOE2q38

Watch

I'm aware of the megahertz myth. At the end of his talk in that video you linked, John Rubinstein said he hopes that people understand that there's more to a processor than its clock speed. The OP is considering 2 processors at the same clock speed, but he's also asking to ignore the differences in cache or bus speed when comparing the two, and still asking if one is faster than the other. I just don't think that makes sense. Again, could be semantics on my part. :p

NC MacGuy
Dec 7, 2008, 11:48 AM
65nm means more heat and more juice needed to push the 1's & 0's. Heat causes resistance which would slow the processes but if given the proper cooling and voltages, i.e. no limitation by enclosure, fans, air flow, etc. they should be theoretically the same. There's a whole world of crap you can look at theoretically speaking but I think that they are very equal in processing speed if just asking that question of the two processors alone.

This is assuming the same architecture. The megahertz myth was assuming two totally different processor architectures.

Stratus Fear
Dec 7, 2008, 06:03 PM
Yes. The cache is double on the new model (6mb vs. 3mb).

Ironically it has a larger cache than all 3 Macbooks, and even the baseline pro.

First model MBA w/ the Merom C2D has 4MB cache, actually.

QCassidy352
Dec 7, 2008, 06:35 PM
just wondering - why would anyone care whether one is faster than the other independent of RAM speed, cache, or bus? It's a purely theoretical question because in the real world, all of those things are inextricably linked to the 2 versions of the MBA. The only question that matters is how one performs compared to the other all things considered because when you're actually using it, all things *will* be considered.

NC MacGuy
Dec 7, 2008, 06:40 PM
^^^^ Exactly. The two processors may be equal in speed but how they're used with all the other pieces parts make the new 1.6GHz a much faster computer.

andyOSX
Dec 7, 2008, 07:41 PM
I'm aware of the megahertz myth. At the end of his talk in that video you linked, John Rubinstein said he hopes that people understand that there's more to a processor than its clock speed. The OP is considering 2 processors at the same clock speed, but he's also asking to ignore the differences in cache or bus speed when comparing the two, and still asking if one is faster than the other. I just don't think that makes sense. Again, could be semantics on my part. :p

No it's not semantics it's a lack of understanding. GHz measure frequency. A true measure of speed would be something like Gigaflops. GHz measure the frequency (speed/cycles) of the processor. A processor with lower frequency but superior architecture is yes running slower but it also has a shorter distance to run. IE that is why the core 2 duo /iTanium etc are all faster than a 3.0GHz P4 at even half the clock speed.

In this case they are the same processor but the more efficient 45nm process applies. Basically there is less physical area that the electrons have to travel around.

This is not "opinion" or "semantics" it is science. Look it up.

andyOSX
Dec 7, 2008, 07:51 PM
I'm aware of the megahertz myth. At the end of his talk in that video you linked, John Rubinstein said he hopes that people understand that there's more to a processor than its clock speed. The OP is considering 2 processors at the same clock speed, but he's also asking to ignore the differences in cache or bus speed when comparing the two, and still asking if one is faster than the other. I just don't think that makes sense. Again, could be semantics on my part. :p

Uh, I don't know if you watched the whole video but actually the key to what he was saying was that in order to reach frequency speeds like 3 GHz the processor makers have to add many pipeline stages. So while data moves faster through the pipeline stages, it has more pipeline stages to go through and thus ultimately it is not necessarily faster.

ashowkati
Dec 7, 2008, 08:26 PM
The clock speed is 1.6 for both. So to answer your question, no. But if you're asking about the processor as a whole, what makes the new processor "the new processor" is all those things (cache, bus, etc.)

It goes even beyond the CPU to fully answer the question. Because of the new nvidia 9400M chip, it takes some of the load off of the CPU. It all ties into performance. Plus if you have the SSD, the rate of which data is accessed is higher, thus giving the illusion of faster "processing."