Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Originally posted by esheep2001
The barefeats results make sad reading :( A comparison of one of the slowest Centrinos agains the fastest G4 and the G4 can barely hold it's head up :(

e.

Forgive me, but you obviously didn't read the entire test result, or you wouldn't have posted such nonsense. Had you read the results and paid attention, you'd have seen that the Centrino beat the G4 in some tests, it was soundly b*tch-slapped by the G4 in the remeining tests.:D
 
Originally posted by esheep2001
The barefeats results make sad reading :( A comparison of one of the slowest Centrinos agains the fastest G4 and the G4 can barely hold it's head up :(

they lost two out of six test. the photoshop one they lost by two seconds and the other is not a real-world test any way. what's to complain about? The G4 is going through that will be sorted out soon enough.
 
Re: Ti vs Al xbench

Originally posted by esheep2001
I posted this on the last thread but many may have finished reading that.

Here is a comparison between a 1GHz/1GB Ram Ti and 1.25GHz/1GB Ram new Al.

The site is in French but if you scroll right down the figures speak for themselves :)

http://www.xrings.net/xrings/article.php3?id_article=239

e.

thanks for that link. That, plus the barefeats results, have sold me: the lack of L3 cache doesn't matter. The new powerbooks soundly beat the old ones. Now if only I could afford one... ;)
 
shadowfax:

L3 cache is pretty useless if you have real DDR RAM. i think the L2 is much more important.
Wowa there! L3 is not useless at all, even with high-speed FSB's and RAM. For example, consider Intel's Xeons with the L3's, and consider their upcoming "extreme" Pentium thats been in the news recently (apparently 90nm is delayed :) ). You can bet the G5 would be much faster (at some things) if it had Xeon-like L3's, as would be the 7457. I'll hazard to guess that with 512k of L2 Apple decided that a 1MB L3 wasn't significant, a 2MB L3 was too expensive, and that there was no need to have an L3 on a older-tech machine than the G5 desktops. Don't want people asking why the G5 doesn't have an L3, do we? ;)

rog:

I guess that's the key question. Do these PBs finally make real use of DDR or is it still a hack job with minimal benefits like on the MDD towers & iMacs?
Pretty sure its the "fake" DDR. The 7447/7457 has always been said to be pin-compatible, which means it uses the same FSB tech. But really, that is not nearly such a problem as everyone tries to make it out to be. Sure, it slows down some super-optimized AltiVec stuff, but 166mhz x 64bit is plenty for just about everything non-AltiVec that a 1.25ghz G4 can do. G4's are just not that fast. (Of course removing the L3 increased demand on the FSB/RAM).

Why don't these "Pro" models have 128MB VRAM, 7200 or at least 5400 RPM drives as standard?
In addition to costing more, these items would generate more heat. I'd say they would make nice options, but not standards.

Some_Big_Spoon:

Does the new 12" have 7455 or 7457? I'm looking at every available doc and can't get the answer.
512k L2 on-die, on a G4 == 7447/7457. 256k L2 on-die, on a G4 == 7445/7455 or 7450 (and 7440?).

csimmons:

Forgive me, but you obviously didn't read the entire test result, or you wouldn't have posted such nonsense.
The P-M was very compeditive with CPU power, but died on the games. Probably a really crappy integrated video chip (I didn't see if they said what it was anywhere).

Edit: Fixed statement about PM performance before someone complains.
 
re: why

Originally posted by Ti_Pousin
Why didn't they put L3 cache? no performance improvement, or cost effective not good enough???

I would guess mostly for heat reasons. It's cooler without that SRAM chip on the motherboard. (Battery lasts longer, too.)

Plus this way Apple can put one on later if they need to rev the AlBooks and the processors aren't enough faster by themselves

Also a 1MB L3 cache won't help a 512KB L2 cache as much is it helps a 256KB L2 cache. The larger the L2 cache is, the larger the L3 cache has to be for the same incremental improvement.
 
Re: 12" Processor?

Originally posted by Some_Big_Spoon
Does the new 12" have 7455 or 7457? I'm looking at every available doc and can't get the answer.

From MacRumors feed during the keynote, it was stated that "Yes, they ALL have the new 7457", but that was the only conformation I've had.

Arn, any word?

Since the 12" PowerBook Processor also has 512k of L2 cache, that means it's either the 7447 or 7457. I would bet 7457 since Apple is using it in their 15, and 17 inch.

It doesn't really matter... since the 7457 and 7447 are functionaly the same. Like I said, the only diff is that the 7457 as the capability to use L3 cache. But none of the PowerBooks have L3 cache... so the difference is academic.

arn
 
Originally posted by csimmons
Forgive me, but you obviously didn't read the entire test result, or you wouldn't have posted such nonsense. Had you read the results and paid attention, you'd have seen that the Centrino beat the G4 in some tests, it was soundly b*tch-slapped by the G4 in the remeining tests.:D

No. I did read the rest of the tests. My point is that this is a comparison of the _SLOWEST_ Centrino against the _FASTEST_ G4. Ok so I guess I should wait for the 1.7 Centrino results (or find some as I'm sure there must be some somewhere) to get a true feeling for just how good/bad this is. The problem is that Intel are on a roll with the Centrino with yet more speed bumps round the corner, where are Moto/Apple? They are at their bleeding edge and struggling to get more out of an ageing and decrepit architecture. I really _really_ hope IBM can pull something out of the hat but at this moment in time it doesn't feel good. Compared to past pbooks yes, but compared to the competition, well, it's like the powermacs were a few months ago without the advantage of MP. Without Apples supporters and fans I feel pbooks would be being left on the shelves in droves :(

e.
 
According to Barefeats, new PowerBooks are slower than Centrino 1.3GHz in Photoshop functions. That sucks. Photoshop should be an area where Macs totally dominates PCs.
 
Re: Re: Re: Re: next update?

Originally posted by esheep2001
Hmm, I've heard similar. I heard simultaneous 3GHz roll out of power mac and books, July 2004 and the fuel cell is actually giving something like 2 days typical use from a single charge/cannister. I also heard that the design will be along the lines of the G5 desktop only the fans will be blowing downwards thereby allowing the laptop to float comfortably in mid air :)

I've always said give the public what they want :)

I'm certainly interested in what magic IBM can create with the G3, especially now that the pro desktop line is sorted with the G5 allowing the G3 to flourish a little. A high-clocking G3 with 'Velocity Engine' (I love Apple marketing, but that sounds so Casio) - the rumoured Mojave chip, possibly in a dual form, could re-invigorate the pb line-up.

On the flip side, now the 970 is here, a lot of people (myself included) think a G5 pb is inevitable. Apple really did make a statement with the G5's case design and heat sinks (scorchio!), but other than the beefed-up G3, what options do they have? Sticking with Motorola can't be a long-term solution.
 
Cripes, bunch of whiners. Apple did about as good as they could at this time with what they had, stop yer grousing. It did pretty well against Centrino in some aspects, admittedly a low end one. I checked around at some other sites like Dell and Compaq and you load up a Centrino speed to outpace a Mac and the prices start to level out all without giving you a solid operating system like OS X not to mention the fit and finish of the new PBs are amazing... fit and finish on a Dell...? Hello, piece a hunka plastic crap!

These aren't gee whiz wow make the peecee users jealous machines... but they're darn good machines. If I had the extra cash I'd be getting one myself.
 
Originally posted by csimmons
Forgive me, but you obviously didn't read the entire test result, or you wouldn't have posted such nonsense. Had you read the results and paid attention, you'd have seen that the Centrino beat the G4 in some tests, it was soundly b*tch-slapped by the G4 in the remeining tests.:D

In the game tests, yes, but which GPU did the centrino use?
 
Originally posted by Photorun
Apple did about as good as they could at this time with what they had, stop yer grousing. It did pretty well against Centrino in some aspects, admittedly a low end one. I checked around at some other sites like Dell and Compaq and you load up a Centrino speed to outpace a Mac and the prices start to level out all without giving you a solid operating system like OS X not to mention the fit and finish of the new PBs are amazing... fit and finish on a Dell...? Hello, piece a hunka plastic crap!

Yep. That is one of the pbooks saving graces but often that doesn't matter to the blinkered Intel/M$ buyer. All they see is speed (MHz) and cost. Style, OS, finish, ease of use are all things that are secondary unfortunately.


These aren't gee whiz wow make the peecee users jealous machines... but they're darn good machines. If I had the extra cash I'd be getting one myself.

Same here, despite my whingeing! I would love a 15" 1.25 but as a second machine, not as my primary, mainly because I'm doing video/audio work for which I will have a G5. Unfortunately, a Centrino looks more attractive as a secondary machine as I don't need it to be fully loaded and I don't need all the bells and whistles but processor speed is an issue. But the main thing is that I can get one for less.

e.
 
What processor did the old 12" use? If it was the 7445 and not the 7455, the 7447 would be a pin compatible drop in replacement and a much easier motherboard 'redesign'.

Lack of L3 seems to make miniscule difference in the benchmarks on the 15/17 but it'd be nice to see some 12" benchmarks as it should now actually be faster in all tests than an iBook.

The Barefeats benchmarks seem a little odd to me. The Centrino 1.3 seems to have a particularly bad graphics chip though perhaps not as bad as the nVidia cards in the old 17"! It's quite possible it would have done better if it had. Not quoting the graphics cards isn't very useful.

The Bryce benchmark is interesting too as it shows a 900Mhz iBook which has a 512K L2, beating a 1Ghz 256K L2 17" G4. I presume Bryce doesn't go well with L3 caches. The P-M 1.3 get's trounced there too which goes to show Cinebench and Photoshop aren't running optimally on the Mac or Bryce on the PC isn't - perhaps Bryce used a different compiler. Motorola also released new software libraries with the 7447 and 7457 with improvements over gcc - are Adobe, Maxxon etc using those?
 
, a Centrino looks more attractive as a secondary machine as I don't need it to be fully loaded and I don't need all the bells and whistles but processor speed is an issue. But the main thing is that I can get one for less.

Buying a PieceCrap and following a pack instead of being original and getting a better overall machine for a little more... no wonder your handle is esheep.
 
Benchmark: Centrino vs. PowerBook

I question those benchmarks. Does anyone know what video card was used in the Centrino laptop? In fact, does anyone know the make or model?

It is not fair to say:

"I threw in the Cetrino 1.3GHz numbers for a reality check. As you can see, it's faster in some ways and slower in others. But the newest PowerBooks are too cool for the competition. Game over"

At 1.3Ghz, that is a very low end Centrino. In general, the price to performance ratio is better for Intel. Even IBM has more cost effective laptop models than Apple.

My complaint has nothing to do with the new PowerBook models. In fact, I am strongly considering 'making the switch'.

But those benchmarks are hardly objective. No consideration is given to price. No specs. are given, etc. My suspicion is that the Centrino would have easily won all of the tests if it had a better graphics card. In any case, I am sure it costs far less than the PowerBook.
 
Originally posted by Websnapx2
they lost two out of six test. the photoshop one they lost by two seconds and the other is not a real-world test any way. what's to complain about? The G4 is going through that will be sorted out soon enough.

I agree -- that was not a real-world test. Furthermore, it is difficult to benchmark across platforms.

However, the video card could have easily been the reason the Centrino lost the other two tests.
 
But those benchmarks are hardly objective. No consideration is given to price. No specs. are given, etc. My suspicion is that the Centrino would have easily won all of the tests if it had a better graphics card. In any case, I am sure it costs far less than the PowerBook.
I don't know about "would have easily won all of the tests" but I do think graphics card was a big factor that resulted in poor performance in lower half of the tests.

I prefer Macs over PC's despite this result because I do believe Macs are higher quality machines over all. But it still is disappointing.
 
Photorun:

Buying a PieceCrap and following a pack instead of being original and getting a better overall machine for a little more... no wonder your handle is esheep.
Your post is eliteism, immaturity and computer-bigotry all wrapped into one bundle. How nice.
 
Originally posted by mdntcallr
Good golly, I love these new powerbooks. only thing is that the technology in them is suited for last June instead of a today release.

Hope the next version has a DVD+-RW Drive.
Better faster mobile graphics chip


they are pretty good for what they are though. I am happy enough i may order one. only thing is that these are pretty much the same machines they should have released months ago.

Bummer

Well, who's to say it isn't a DVD±RW drive? Mike at XLR8YourMac had never come across the Superdrive used in one of his reader's 15" ones, and that's saying something. Admittedly, it's probably disabled until Panther if it is, but still, perhaps better than nothing.
 
Originally posted by esheep2001
No. I did read the rest of the tests. My point is that this is a comparison of the _SLOWEST_ Centrino against the _FASTEST_ G4. Ok so I guess I should wait for the 1.7 Centrino results (or find some as I'm sure there must be some somewhere) to get a true feeling for just how good/bad this is. The problem is that Intel are on a roll with the Centrino with yet more speed bumps round the corner, where are Moto/Apple? They are at their bleeding edge and struggling to get more out of an ageing and decrepit architecture. I really _really_ hope IBM can pull something out of the hat but at this moment in time it doesn't feel good. Compared to past pbooks yes, but compared to the competition, well, it's like the powermacs were a few months ago without the advantage of MP. Without Apples supporters and fans I feel pbooks would be being left on the shelves in droves :(

e.

you're basically right about this. The G4 is a dead chip, as far as personal computing is concerned. Apple is just trying to make something out of nothing. The fact that it's even remotely competitive is encouraging. Also, Panther should speed things up, mainly for G5s but also for all other systems, according to early testers. Besides the CPU, the new PowerBooks are seriously tricked out systems, a solid package.
 
The new pbooks are running 10.2.7, right? Any developers out there with insight into how much of the speed diff. (if any) was due to os improvements? Any chance there are improvements to the opengl drivers for the ati cards in the tibook?
 
Originally posted by Photorun
Buying a PieceCrap and following a pack instead of being original and getting a better overall machine for a little more... no wonder your handle is esheep.


So if he buys a machine like those used by tens or hundreds of millions of others, he's a 'sheep'. But if he buys one used by only millions of others, he's being 'original'.


Strident nonsense like this is exactly the kind of advocacy the Mac doesn't need. It's people like you that get Mac users branded as kooks or elitists.


Anyone who sees you arguing like this about platforms is going to conclude that you advocate Macs, not because you've made some sort of honest or objective comparison of the relative merits of the platforms, but to rather to fill some psychological need you have to differentiate yourself from 'the herd'.
 
Originally posted by shadowfax
boy, won't it be nice when they get 1 MB L2 Cache...

L3 cache is pretty useless if you have real DDR RAM. i think the L2 is much more important.

Each "level" is approximately an order of magnitude slower than the last on typical (ie, non-burst) accesses. While the L3 isn't much faster than main memory on a clock-for-clock basis, you must take into consideration:

1) L3 doesn't lie on the other side of the System Controller and CPU FSB. Accessing L3 is significantly faster than the FSB, even without taking latencies into account.

2) L3 is very low-latency. You don't waitn 20+ cycles to pull data from L3; it is just a few cycles.

IOW, fast DDR memory is certainly no replacement for L3 cache.

More L2 cache will help significantly, but it just expands the "hot zone size"; if your app uses more than 512K data then you're stuck going to 10-100x slower RAM.

The theory being, of course, that most apps' hot zone will fit inside L2, or at least a significant enough portion that L3 isn't worth the extra cost.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.