PowerBook Updates in September?

[Squire]

So for a while Apple is going to sell iBooks that are cheaper than PowerBooks and that blow them away speed-wise? And iMacs that are cheaper than powermacs but twice as fast? I don't see that happening. They'll convert the high-end stuff to Intel first, not the low end, though I'll admit that the Mac Mini could be an exception to that rule, as it would bring in switchers.

If you're responding to my comments in the post above, perhaps you should re-read what I wrote. People keep on saying over and over that the intel chips are going to go in the low-end models first. I said that I don't necessarily agree with that assumption. I specifically stated that I thought Powerbooks would be one of the first. Perhaps I should have made my views clearer.

The way I see it, engineering factors will determine which models get the intel chips first. Therefore, the Powerbook G4 is a prime candidate in my book. It probably can't go much higher without significant internal restructuring. The PowerMac models, on the other, might see dual-core G5 chips. The other, consumer models still have some room to grow.

Squire

<edit> Good point, gekko513. I doubt that there will be very many machines that could trounce a dual-core, dual processor G5 PowerMac should one ever become available.

 

[minimax]

Sorry, which link in that post showed evidence that the Macintel "beat the snot of of the dual 2.7"? I checked them all and failed to find any such quote. In fact, the final link in that post showed that xBench scores were only about 1/3 the speed on a Macintel as they were on a dual 2.7.

The evidence is that the OS X UI 'felt snappier' and the beach ball was spotted less than on a dual G5 silly. How much more proove do you need that PPC is wack?

 

[minimax]

Or, could this be the reason:

And I think if anything can be made from these reports, it's that when OSX is running intel optimized graphics drivers, the visible impression of speed - "Teh Snappy" - is more evident. This is because the majority of the work that goes into producing graphics drivers is to get optimal speed out of the Intel architecture, not PowerPC.

Otherwise it would be extremely hard to believe that somehow code running on OSX, compiled in GCC 4.0 for Intel has suddenly opened up a lead where a single Pentium 4 is running quicker than a Dual G5. Every single benchmark in the last few years has indicated you need a Dual Opteron or Xeon to get close to the G5 PowerMacs. And that's usually in benchmark wars with Windows where the MS and Intel compiler is a lot more efficient than GCC.

My curry is spicier, thus, my tomatoes 0wn yours.

 

[AidenShaw]

Sorry, which link in that post showed evidence that the Macintel "beat the snot of of the dual 2.7"? I checked them all and failed to find any such quote..

Note that my post said "what are they saying" about the MacIntel. The "snot" remark is just one of many comments that have been posted about the good performance of the MacIntel developer systems.

Use the "find text" feature of your browser and search for "snot" on that page (http://www.neowin.net/forum/lofiversion/index.php/t330170.html), you should see:

Jack31081 Jun 9 2005, 13:57
The more I read about the Intel switch, the more i think Apple is making an awesome move.

A Mac that will run Windows natively and possibly get a bunch more game ports due to easier transition from Windows code to Mac code. And the fact that the Intel 3.6 seems to beat the snot out of the G5 2.7. I mean, hell, Tiger runs great on my 1Ghz G4...I can only imagine what it'd be like running it on a 4Ghz, dual-core pentium.

In fact, the final link in that post showed that xBench scores were only about 1/3 the speed on a Macintel as they were on a dual 2.7.

And the first sentence of that final link says "benchmark reports of programs running under Rosetta"....

Xbench was emulated/translated....and still outperformed the dual 2.5 in graphics.
______________________

Note that the Steve really doesn't want benchmarks done on the MacIntel developer machines...and probably has the lawyers ready to pounce in case reports show up.

We won't see good benchmarks until Apple starts officially selling MacIntels.

Until then, however, the following is interesting: SPEC Scores

  Model       Speed    CPU Type   SPECint   SPECfp
----------  --------   --------   -------   ------
IBM JS20     2.2 GHz    PPC970       986     1178
IBM HS20     3.6 GHz     Xeon       1684     1769

Is it fair to say that the Xeon "kicks the snot" out of the PPC970 here?

Wouldn't it be reasonable to believe that on applications that don't use MP well, the Xeon would "kick the snot" out of a dual 2.7?

 

[gekko513]

The evidence is that the OS X UI 'felt snappier' and the beach ball was spotted less than on a dual G5 silly. How much more proove do you need that PPC is wack?

The GUI felt snappier and the beach ball was spotted less ... doesn't sound scientific to me ... sounds like some of the forum posts that claim Safari felt snappier after every single 10.3.X update no matter what had actually been updated. (Edit: OK, the rolleyes was there for a reason )

 

[fordlemon]

It's pretty clear that they were lying about the "PowerPC Rules - look at these 'benchmarks'"....

The single processor P4 MacIntels are "kicking the snot out of dual 2.7 G5s" according to some reports on the web.

But, you expect lies in adverts, right?

No, Apple has lied to us about the superiority of their hardware over the years. They advertised that ghz is a myth. That the PPC cpu was so much better at crunching big chunks of data and is why they are better for video and audio editing over any x86 based machine. You buy an Apple because of the hardware. Now they no longer have any leverage. They will be just like any other pc, but they made a huge mistake going with Intel rather than AMD. Stability in an OS is no longer an issue. Why are they going backward in technology? x86 is an old outdated platform. Apple no longer cares about their loyal customer base. It's all about greed. A pure business decision. It's too bad. I'm sure Woz isn't happy about it.

Here's an excerpt from a Slashdot interview with Woz about why Apple hardware is better then x86.

"Many of the hardware advantages that Apple has is due to it's being more tightly controlled by Apple and in it's being more tightly integrated with the software. That allows Apple to make hardware changes and decisions that are more reliable than in the Wintel world. This has nothing to do with Linux and everything to do with MacOS. The basic plumbing is superior to Intel hardware in some ways (firewire on the motherboard for example) and a bit lacking in others (3D rendering hardware) but the basic performance advantage goes to the RISC architecture of the PowerPC processor. Intel's response to this is that even if RISC is 40% faster, that only amounts to a few months lead, according to Moore's Law."
http://slashdot.org/article.pl?sid=00/01/07/1124211&mode=thread

 

[minimax]

No, Apple has lied to us about the superiority of their hardware over the years. They advertised that ghz is a myth. That the PPC cpu was so much better at crunching big chunks of data and is why they are better for video and audio editing over any x86 based machine. You buy an Apple because of the hardware. Now they no longer have any leverage. They will be just like any other pc, but they made a huge mistake going with Intel rather than AMD. Stability in an OS is no longer an issue. Why are they going backward in technology? x86 is an old outdated platform. Apple no longer cares about their loyal customer base. It's all about greed. A pure business decision. It's too bad. I'm sure Woz isn't happy about it.

yeah yeah whatever.

@AidenShaw: thanks for that SPEC benchmark. BUT it is not fair to compare a 2.2 G5 with a 3.6 P4. Of course the P4 will be faster. It would also not be fair to denounce the G5 on the basis of one specific SPEC benchmark (which btw nowhere mentions the PPC platform so i can only hope it has been optimised for that type of architecture as well) and only ONE IBM server that runs the G5.

There is only one decent comparison between the P4 XEON, the G5 and the Opteron which shows CLEARLY that the 2.7 G5 is a match for the P4 3.6. http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=5

 

[fordlemon]

yeah yeah whatever.

@AidenShaw: thanks for that SPEC benchmark. BUT it is not fair to compare a 2.2 G5 with a 3.6 P4. Of course the P4 will be faster. It would also not be fair to denounce the G5 on the basis of one specific SPEC benchmark (which btw nowhere mentions the PPC platform so i can only hope it has been optimised for that type of architecture as well) and only ONE IBM server that runs the G5.

There is only one decent comparison between the P4 XEON, the G5 and the Opteron which shows CLEARLY that the 2.7 G5 is a match for the P4 3.6. http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=5

You don't get it. Not only is it the cpu it is the rest of the hardware architecture. Also, how the OS and other Apple software is written to use specific instruction sets in the PPC cpu. The new Apples will not have this to their advantage, making them just another slave to Intel. I say this and I still have my Apple II. It's a sad time for loyal Apple users. They are no longer better or different, just another Wintel box.

 

[minimax]

You don't get it. Not only is it the cpu it is the rest of the hardware architecture. Also, how the OS and other Apple software is written to use specific instruction sets in the PPC cpu. The new Apples will not have this to their advantage, making them just another slave to Intel. I say this and I still have my Apple II. It's a sad time for loyal Apple users. They are no longer better or different, just another Wintel box.

Yes i don't get you. Either you are trolling or you should try to get your thinking straight. Now, you say you buy a Mac because of the hardware, and with the switch there wont be this advantage anymore. On the other hand Apple lied about the superiority of the hardware. Now which of both is it exactly?

And you say Apple can't write for the instruction sets of x86? haha they can pack their business in that case.
My advise: just stick to your Apple II my friend, and leave the technical aspects of OSx86 to Apple.

 

[devastator777]

maybe but i don't care, cause i love my current PB!!!

DITTOOOOO...


Dual 867 22" ACD, 17"ACD, 17" 1.5 ghtz PB, 60 gb ipod, 20 1st gen IPOD.

 

[AidenShaw]

@AidenShaw: thanks for that SPEC benchmark. BUT it is not fair to compare a 2.2 G5 with a 3.6 P4. Of course the P4 will be faster.

But, MHz is a myth - so why is it unfair?

Note that the 2.2 GHz PPC970 is about 80% of the clock rate of the fastest PowerMac. If we scale the SPECint numbers by that ratio (being generous and assuming linear scaling), we get

  Model       Speed    CPU Type   SPECint   SPECfp
----------  --------   --------   -------   ------
IBM JS20     2.2 GHz    PPC970       986     1178
IBM JS20     2.7 GHz    PPC970      1210     1445  (est)
IBM HS20     3.6 GHz     Xeon       1684     1769

So, even being generous to the PPC970, it still gets kicked by the P4.

It would also not be fair to denounce the G5 on the basis of one specific SPEC benchmark (which btw nowhere mentions the PPC platform so i can only hope it has been optimised for that type of architecture as well) and only ONE IBM server that runs the G5.

I'm not "denouncing" the PPC970. I'm merely saying that there are reports on the web that the MacIntel is a very fast machine, often faster than the G5.

Looking at the SPEC numbers, it does seem reasonable to conclude that perhaps this is true for many apps (in particular, those that don't exploit dual processors and those where SSE is a match for AltiVec).

And, it's not my fault that IBM only uses the PPC970 in one single server in their entire lineup!

There is only one decent comparison between the P4 XEON, the G5 and the Opteron which shows CLEARLY that the 2.7 G5 is a match for the P4 3.6. http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=5

"Decent" - because you like the results? The text of that article makes it pretty clear that the gcc compiler used is pretty poor at generating good code. (Nothing new here....)

Also, that section is focussed on floating point. Most user apps are primarily integer-based - where the P4 has a bigger advantage.
___________________


But, I don't need to convince you that Intel's chips are better. Intel convinced Jobs, and you're going to be getting them.

 

[minimax]

But, MHz is a myth - so why is it unfair?

Note that the 2.2 GHz PPC970 is about 80% of the clock rate of the fastest PowerMac. If we scale the SPECint numbers by that ratio (being generous and assuming linear scaling), we get

  Model       Speed    CPU Type   SPECint   SPECfp
----------  --------   --------   -------   ------
IBM JS20     2.2 GHz    PPC970       986     1178
IBM JS20     2.7 GHz    PPC970      1210     1445  (est)
IBM HS20     3.6 GHz     Xeon       1684     1769

So, even being generous to the PPC970, it still gets kicked by the P4.


I'm not "denouncing" the PPC970. I'm merely saying that there are reports on the web that the MacIntel is a very fast machine, often faster than the G5.

Looking at the SPEC numbers, it does seem reasonable to conclude that perhaps this is true for many apps (in particular, those that don't exploit dual processors and those where SSE is a match for AltiVec).

And, it's not my fault that IBM only uses the PPC970 in one single server in their entire lineup!

No it's not your fault but it doesnt make the benchmark very reliable does it?
Also, I'm not an expert, but somewhere the SPEC benchmark was said to be platform independant, but from all platforms mentioned (they were quite a few) PPC was lacking.

"Decent" - because you like the results? The text of that article makes it pretty clear that the gcc compiler used is pretty poor at generating good code. (Nothing new here....)

Also, that section is focussed on floating point. Most user apps are primarily integer-based - where the P4 has a bigger advantage.

You can't deny that also with all other tests (CINEBENCH/ Lightwave / Povray) there were similar results that clearly show the performance matches up. This test has been far more elaborate then any of the tests you or others have been able to come up with (like that shady barefeats one) so I could turn that question about my insinuated subjectiveness just as easy around. You're not exactly in the position to make such accusations, as I havent seen you show a similar extensive benchmarkt to proove this one wrong.

And regarding those reports on the web. Perhaps scroll some posts back...I think you missed them. Not only are they as subjective as can be, but they don't take into account the many differences between both platforms (bios, video drivers, hard disk performance etc. etc.) You are not exactly building your case any stronger by insisting on mentioning those reports again.

 

[AidenShaw]

Also, I'm not an expert, but somewhere the SPEC benchmark was said to be platform independant, but from all platforms mentioned (they were quite a few) PPC was lacking.

There are only two users of the PPC970 architecture making commercial systems.

IBM only uses the PPC970 (@2.2 GHz) in its JS20 blade system. That's what was in the SPEC tables. It's interesting, since IBM also uses Xeon chips in the same blade system chassis. This means that the numbers are for comparable systems (both are two-way blades).

The other company is Apple, which works hard to avoid reporting cross-platform standard benchmarks. Apple doesn't submit to SPEC.

You can't deny that also with all other tests (CINEBENCH/ Lightwave / Povray) there were similar results that clearly show the performance matches up.

I never denied this, and was careful to always mention that MP-aware and "AltiVec sweetheart" apps were not being discussed.

And regarding those reports on the web. Perhaps scroll some posts back...I think you missed them. Not only are they as subjective as can be, but they don't take into account the many differences between both platforms (bios, video drivers, hard disk performance etc. etc.) You are not exactly building your case any stronger by insisting on mentioning those reports again.

Please reread my posts - I've always treated these as anectdotal evidence. The buzz on the web is that the MacIntels are wicked fast. That's true - about the buzz. The subjective experience of many people shouldn't be dismissed.

Does it matter why - I keep hearing that it's the "whole platform" that matters, not any one component. The "whole platform" of the MacIntel is reported to be speedy.

I'm sure you all will be thrilled when Apple introduces the first 4 processor MacIntels (dual dual-core 64-bit Xeon chips) - and you'll agree that it kicks the snot out of the PPC Macs.
_____________________

Please also note that the original discussion was about shared graphics - and how shared graphics cripples a system.

The MacIntel developer system has Intel shared graphics, and certainly doesn't seem to be crippled according to the stories on the web. The argument wasn't PPC vs Pentium, but to try to rebut the "shared" == "crippled" claim.

You've turned it into PPC vs Pentium, so let's call a truce here....

 

[iQuit]

There is just something about PPC that makes it go with a Mac...like milk and cookies or a doughnut and coffee.


http://www.anandtech.com/mac/showdoc.aspx?i=2436&p=2

 

[minimax]

_____________________

Please also note that the original discussion was about shared graphics - and how shared graphics cripples a system.

The MacIntel developer system has Intel shared graphics, and certainly doesn't seem to be crippled according to the stories on the web. The argument wasn't PPC vs Pentium, but to try to rebut the "shared" == "crippled" claim.

You've turned it into PPC vs Pentium, so let's call a truce here....

Fair enough, let's wait for the first native x86 applications for the Macintel before we draw any conclusions

 

[fordlemon]

Yes i don't get you. Either you are trolling or you should try to get your thinking straight. Now, you say you buy a Mac because of the hardware, and with the switch there wont be this advantage anymore. On the other hand Apple lied about the superiority of the hardware. Now which of both is it exactly?

And you say Apple can't write for the instruction sets of x86? haha they can pack their business in that case.
My advise: just stick to your Apple II my friend, and leave the technical aspects of OSx86 to Apple.

You don't get it at all. You have no idea how the current Apple hardware works. I run a protools rig that runs circles around anything an Intel platform can put out because of Apples RISC platform. It will no longer be superior. People like you that buy them just because they are pretty and can not comprehend the technical advantages will still buy them.

 

[fordlemon]

Why buy a PPC PB if the Macintel will be cheaper and better?

Why would anybody buy an Apple now if you all believe that the Macintel will be better and cheaper? After all, one reason for the switch is to lower prices. It's too bad AMD couldn't have provided a platform for Apple like Intel can. If they must switch to the x86 architecture that they have been bashing for decades, I wish they could've used the better cpu.

 

[wdlove]

Why would anybody buy an Apple now if you all believe that the Macintel will be better and cheaper? After all, one reason for the switch is to lower prices. It's too bad AMD couldn't have provided a platform for Apple like Intel can. If they must switch to the x86 architecture that they have been bashing for decades, I wish they could've used the better cpu.

The PPC still does a good job for most users. Also many prefer not to purchase a Rev. A model Mac, let others be a beta tester.

 

[Platform]

Why would anybody buy an Apple now if you all believe that the Macintel will be better and cheaper? After all, one reason for the switch is to lower prices. It's too bad AMD couldn't have provided a platform for Apple like Intel can. If they must switch to the x86 architecture that they have been bashing for decades, I wish they could've used the better cpu.

Because some people don't like Intel and simply will pay more just not to have an Intel.......

Because they need a laptop now and can't wait a year or more.....

Because people are scared of Rev 1. especially since new CPU form their previous enemy....ETC.

 

[Surreal]

Why would anybody buy an Apple now if you all believe that the Macintel will be better and cheaper? After all, one reason for the switch is to lower prices. It's too bad AMD couldn't have provided a platform for Apple like Intel can. If they must switch to the x86 architecture that they have been bashing for decades, I wish they could've used the better cpu.

who said the macintels will be cheaper? or even all the way better initially? i'm excited for architecture advancements like pcix or e or whatever, DDR2-3 and FSB increases, but i could be made to be horriby wrong if those dot equal a better computing experience.

i REALLY want a freaking G5 in a laptop. improbable as it is. floating point calculations and front side bus = lovely audio processing machine.

 

[minimax]

You don't get it at all. You have no idea how the current Apple hardware works. I run a protools rig that runs circles around anything an Intel platform can put out because of Apples RISC platform. It will no longer be superior. People like you that buy them just because they are pretty and can not comprehend the technical advantages will still buy them.

OK Einstein if you say so.

 

[gekko513]

Also, that section is focussed on floating point. Most user apps are primarily integer-based - where the P4 has a bigger advantage.
___________________


But, I don't need to convince you that Intel's chips are better. Intel convinced Jobs, and you're going to be getting them.

I've been doing some Cocoa programming for the last year and before I started that, I was also of the opinion that most user apps are primarily integer-based. With Cocoa that isn't true as many GUI values are floats. Screen coordinates are floats and colors are 4 floats. This would seem to indicate that GUI snappiness depends on floating point performance as well as integer performance. It is, however, hard for me to give an opinion on which is more important, since I don't know enough about the inner workings of Cocoa.

 

[digitalbiker]

Cheaper?

Why would anybody buy an Apple now if you all believe that the Macintel will be better and cheaper? After all, one reason for the switch is to lower prices. It's too bad AMD couldn't have provided a platform for Apple like Intel can. If they must switch to the x86 architecture that they have been bashing for decades, I wish they could've used the better cpu.

If you are waiting for a cheaper PB than right now, I think you will be waiting for quite awhile. Nowhere has Apple said anything about cheaper. The higher end Pentium M chips, Yonah, Merom, etc. are all going to be very expensive. Much more than the G4's.

I am one of those that is cautiously optimistic about the Intel switch.

I would like the possiblity of running native window apps in a shell. I would like the access to all those cutting edge GPU's, HD's, ram, etc. instead of waiting for the mac version to be released.

But I hate the thought of going CISC architecture when RISC is so much better for many scientific endeavers that I am involved with.

I am also very skeptical of Intel, and their ability to innovate, improve, and meet projected roadmaps.

I also wonder just where Apple will sit on the pecking order for access to Intel chips. Will Dell, HP, Sony all get preference for new chipsets that are in short supply, leaving Apple holding the ***ty end of the stick?

I guess we will see!

 

[SiliconAddict]

They are no longer better or different, just another Wintel box.

God. Its like talking to a six year old. You just keep banging your head into the wall in frusteration.

It will NOT be another Wintel box. Why. Damn simple:

1. It runs OS X. No other system without some major hacking will run OS X once. Will some hacker be able to make it run on some other system? Sure but you are going to have to jump through hoops to get it going which will account for may 5% of the users. So the original statement still stands these systems run OS X.

2. These systems will run the iApps Suite of tools. No one else has these.

3. Apple will control the entire creating from start to finish. From OS to CPU. Wintel boxes have no say in the matter when it comes to the OS. MS takes input (Once every 4 years or so) but its hardly the same as having the manufacturer of the hardware custom roll their OS to make it sing on the system.

4. Apple design. Enough said.


Funny how these are the same advantages before Apple moved to x86.
Only the seriously ignorant is going to not see how this is going to strengthen Apple in the long run. Only the Apple zealots of the world give a crap what chip is inside the system. The average user and even the not so average user would never even KNOW that it isn't a PPC if not for the properties box or ripping off the heatsink. A computer is not just the sum of its hardware. Nor is it the sum of the apps and OS installed on it. Its the total package and at the end of the day or more accurately at the end of 2006 Apple will have a kick butt product line that has, debatably, superior hardware, superior design, superior software, and a superior OS.

 

[CubaTBird]

God. Its like talking to a six year old. You just keep banging your head into the wall in frusteration.

It will NOT be another Wintel box. Why. Damn simple:

1. It runs OS X. No other system without some major hacking will run OS X once. Will some hacker be able to make it run on some other system? Sure but you are going to have to jump through hoops to get it going which will account for may 5% of the users. So the original statement still stands these systems run OS X.

2. These systems will run the iApps Suite of tools. No one else has these.

3. Apple will control the entire creating from start to finish. From OS to CPU. Wintel boxes have no say in the matter when it comes to the OS. MS takes input (Once every 4 years or so) but its hardly the same as having the manufacturer of the hardware custom roll their OS to make it sing on the system.

4. Apple design. Enough said.


Funny how these are the same advantages before Apple moved to x86.
Only the seriously ignorant is going to not see how this is going to strengthen Apple in the long run. Only the Apple zealots of the world give a crap what chip is inside the system. The average user and even the not so average user would never even KNOW that it isn't a PPC if not for the properties box or ripping off the heatsink. A computer is not just the sum of its hardware. Nor is it the sum of the apps and OS installed on it. Its the total package and at the end of the day or more accurately at the end of 2006 Apple will have a kick butt product line that has, debatably, superior hardware, superior design, superior software, and a superior OS.

just hopefully not a superior price