Apple switching to pentium?

[rutt]

This rumor is all over the internet. It's been brewing for a good long while. Apple really does have a huge preformance problem with PowerPC and looking forward this is only likely to get worse. But the transition seems pretty darn challenging to me. Knowing that Apple is a year or so from introducing much faster products that will obsolete the software investment in PPC, well, does that make you more or less likely to make a new Apple hardware investment?

My horizon on this was a little short, but I think I had the big picture right.

[patch29]

I was going to post this too. The last rumor statements are pretty bold and are coming from cnet, article here. (http://news.com.com/Apple+to+ditch+IBM%2C+switch+to+Intel+chips/2100-1006_3-5731398.html?part=rss&tag=5731398&subj=news)

I don't know about a new purchase. I am hoping they will give us something great tomorrow, but no real rumors, but I still hope (New Powerbook, with new chip, fingers crossed). If not I think I will finally take the plunge on a new G5 dual. It all depends on what they announce tomorrow. I could hold off, but several times a week I need the speed and simply deal with it for now.

[rutt]

I was going to post this too. The last rumor statements are pretty bold and are coming from cnet, article here. (http://news.com.com/Apple+to+ditch+IBM%2C+switch+to+Intel+chips/2100-1006_3-5731398.html?part=rss&tag=5731398&subj=news)

I don't know about a new purchase. I am hoping they will give us something great tomorrow, but no real rumors, but I still hope (New Powerbook, with new chip, fingers crossed). If not I think I will finally take the plunge on a new G5 dual. It all depends on what they announce tomorrow. I could hold off, but several times a week I need the speed and simply deal with it for now.

My guess is that you are unlikely to get that new Powerbook. Notebook chips is one of the big motivators for Apple switching. There just isn't a compettive PPC option for them to use in a notebook; Intel is too far out ahead in this area.

[Richard]

...If not I think I will finally take the plunge on a new G5 dual. It all depends on what they announce tomorrow. I could hold off, but several times a week I need the speed and simply deal with it for now. I would think more than twice before purchasing anything. Long ago, I bought the last Mac based on the 68xxx chip. One year later, Apple introduced the Power PC chip and guess what? Within another year, there was no new software for the lowly 68040 Quadra. Unless you are comfortable that you have all the software you are going to need within the lifetime of your next machine, buying a PowerPC based Mac now seems like a risky investment.

[Andy]

If not I think I will finally take the plunge on a new G5 dual.


oh sure. but y'all have no problem pushing me into one, eh :lol3

[rutt]

oh sure. but y'all have no problem pushing me into one, eh :lol3

Andy, I think you'll live. Me, too. (And I can't even say that I didn't listen to me! I knew what I was doing.)

But what about Apple? How do they manage this transition? What happens to their computer sales over the next year or so? How do they get Adobe and others on board? I suppose, one possibility, is that their pentium based computers actually run windows based software. This isn't that hard a technical challenge, especially if Apple gets a little help from Microsoft, one of its big investors. On linux, the windows emulator, WINE, does an OK but not great job of running PS7. Also, expect Vmware and/or similar on the new Apple platform. Being able to run all the windows software will be a plus for Apple. Competing on a level playing field in terms of processor performance and price/performance will also be a plus. But without native versions of flagship applications, I think the platform will lose differentiation to both windows and linux. So there is an interesting challenge for them.

Don't get me wrong. Steve Jobs is much better at this than I could ever hope to be. I just wish I could read his mind.

[pathfinder]

Andy, I think you'll live. Me, too. (And I can't even say that I didn't listen to me! I knew what I was doing.)

But what about Apple? How do they manage this transition? What happens to their computer sales over the next year or so? How do they get Adobe and others on board? I suppose, one possibility, is that their pentium based computers actually run windows based software. This isn't that hard a technical challenge, especially if Apple gets a little help from Microsoft, one of its big investors. On linux, the windows emulator, WINE, does an OK but not great job of running PS7. Also, expect Vmware and/or similar on the new Apple platform. Being able to run all the windows software will be a plus for Apple. Competing on a level playing field in terms of processor performance and price/performance will also be a plus. But without native versions of flagship applications, I think the platform will lose differentiation to both windows and linux. So there is an interesting challenge for them.

Don't get me wrong. Steve Jobs is much better at this than I could ever hope to be. I just wish I could read his mind.

Sounds like their stockholders are not going to be happy Monday either! Interesting situation - surely this was all discussed in committee before Apple made this decision. This creates uncertainty, and stockholders do not like that! Arghhh.

[DavidTO]

This rumor has been around a LONG time. The only difference this time is that the news source is more reliable.

There are so many factors that we don't know about, it's hard to say how much validity this rumor has.

The biggest problem, IMO, is that the software developers have to go through yet another migration to another platform, and we'll certainly lose some along the way.

OSX is easily ported to Intel, and most likely already is. The Maklar project has been going on for a while, if the rumor sites are at all right about that.

Personally, I think there are other areas that Apple's been slow on. Graphics cards and using them for rendering is something that Core Image is just starting to address.

Anyway, high-end workstations based on Intel chips are most likely 2 years away, 1 year at best, so IMO any investment you've made to date is fine. Heck, at this point Apple could conceivably make this transition before MS even starts to catch up with the release of Longhorn.

And yes, Apple could benefit from faster chips, but the old saw about usability is very valid. Reference the Intel exec who said he spends an hour every week on his daughter's PC dealing with viruses and spyware, and saying that the only way to really deal with it is to get a Mac.This series (http://securityawareness.blogspot.com/2005/05/mad-as-hell-iii-month-1-review.html) is also interesting.

[tlittleton]

I don't know. At first I thought that they wouldn't do it. They just released Tiger, which though I havent used it yet, I understand to be a substantial upgrade(?). But then again, OS-X is a UNIX based OS. I wouldn't think it would be that hard to make some kernel changes and release an Intel x86 compatible version of OS-X.

I would think the main things marketing wise that would be holding Apple back would be that famous "Intel Inside" logo right below the Apple on a G5...

I also think that if they do go to an Intel chip, the chip will not be called a Pentium.

[rutt]

This rumor has been around a LONG time. The only difference this time is that the news source is more reliable.

There are so many factors that we don't know about, it's hard to say how much validity this rumor has.


A lot more reliable. The New York Times and The San Jose Mercury News both have the story on their web sites. Something is going to happen. We just don't exactly know what.


The biggest problem, IMO, is that the software developers have to go through yet another migration to another platform, and we'll certainly lose some along the way.


That's the interesting puzzle. How do they get through this and not end up with some real hard years. I can think of all kinds of things, but the only ones that don't seem like business disasters involve Intel making something that runs PPC instructions. But just how likely is that? I suppose from Intel's point of view, they don't care. Their whole goal is just to keep their fabs busy. The micro architecture underneath modern Pentiums (Yamhill) could just as easily support PPC as x86. The translation engine from PPC to the microcode would be tons simpler than the X86 engine. I believe Apple owns enough PPC architecture for this to be legally feasible. But this is just speculation, of course. I find that an interesting game. We should be able to score it tomorrow.

Personally, I think there are other areas that Apple's been slow on. Graphics cards and using them for rendering is something that Core Image is just starting to address.
But Apple already has world class partners here. They can use anything availble for the PC market. It's the darn processor where they bet on the wrong horse. And as time goes by, the results of this bad bet will just get worse and worse. Say what you like, Apple can't afford to be 1/2 as fast per processor as commodity PCs that are much cheaper. Apple can't afford to compete with multiple core chips when they don't have any such thing. There is a limit to the power of the Jobs Reality Distortion Field.

[DavidTO]

A lot more reliable. The New York Times and The San Jose Mercury News both have the story on their web sites. Something is going to happen. We just don't exactly know what.

Well, they have the exact story from the original news source, only reprinted. Yes, it's in the NYT, but they didn't do any original reporting to back up the other source, they just decided to trust that one original source. We'll see tomorrow.

But Apple already has world class partners here. They can use anything availble for the PC market. It's the darn processor where they bet on the wrong horse. And as time goes by, the results of this bad bet will just get worse and worse. Say what you like, Apple can't afford to be 1/2 as fast per processor as commodity PCs that are much cheaper. Apple can't afford to compete with multiple core chips when they don't have any such thing. There is a limit to the power of the Jobs Reality Distortion Field.

They can use any graphics card, they just don't, and they don't use them as effectively as they could.

Also, I'm not so sure that they bet on the wrong horse. Meaning (and I'm no expert), I don't believe that the technology is inherently any worse, it's just that without the volume to drive it, the progress is going to be slower. The competition for a larger pool of dollars is what is pushing the x86.

[rutt]

Also, I'm not so sure that they bet on the wrong horse. Meaning (and I'm no expert), I don't believe that the technology is inherently any worse, it's just that without the volume to drive it, the progress is going to be slower. The competition for a larger pool of dollars is what is pushing the x86.

It doesn't matter why the horse loses the race. If it doesn't cross the line first, no roses.

PPC is a much nicer architecture than X86. So are MIPS and Sparc for that matter. The nicest of all was Alpha. No matter. Volume wins in today's world. Even Intel can't compete with itself. XA64 (Itanium) is quite a bit slower than Xeon no matter how much Intel wishes this weren't true. X86 volume means it makes sense for Intel to spend virtually anything on each successive X86 design. This just isn't true of anything else.

[DavidTO]

Here's (http://www.eweek.com/article2/0,1759,1824230,00.asp) an interesting article that summarizes the reaction to the rumors.

[DavidTO]

It doesn't matter why the horse loses the race. If it doesn't cross the line first, no roses.

PPC is a much nicer architecture than X86. So are MIPS and Sparc for that matter. The nicest of all was Alpha. No matter. Volume wins in today's world. Even Intel can't compete with itself. XA64 (Itanium) is quite a bit slower than Xeon no matter how much Intel wishes this weren't true. X86 volume means it makes sense for Intel to spend virtually anything on each successive X86 design. This just isn't true of anything else.

Exactly.

[mercphoto]

But Apple already has world class partners here. They can use anything availble for the PC market. It's the darn processor where they bet on the wrong horse. And as time goes by, the results of this bad bet will just get worse and worse.

Being someone who used to work at the Somerset Design Center, part of the AIM alliance that developed the G3 and G4 processors, I have some things I could say about that but won't. Let me just simply say that Apple's processors woes rest as much on their shoulders as it does on IBM or Freescale's. Things look different from the inside.

As per what will happen Monday, while it might involve Intel in some way, I would actually expect more of a new product announcement, probably a home entertainment device, rather than "we're switching processors". I expect the current Macs to stay Power PC, and to see a brand new device on an Intel processor. Just my guessing.

As per the wrong horse theory, I now work for AMD. I've been in both worlds. Texas Instruments for PC's, then Motorola for PPC, now AMD for x86 again. Without a doubt, its the market that bet on the wrong processor architecture, not Apple.

[rutt]

As per the wrong horse theory, I now work for AMD. I've been in both worlds. Texas Instruments for PC's, then Motorola for PPC, now AMD for x86 again. Without a doubt, its the market that bet on the wrong processor architecture, not Apple.

I think that's an oxymoron.

I worked at SGI on two MIPS processors. One of my really good friends was the architect of Pentium 2, so I guess we can blame a lot on him.

X86 is really ugly. But the market bet on it, and it was a self fulfilling prophecy. After Pentium 2, the architectural disadvantage of X86 vs the RISC machine was gone. So the X86 world never had to face the wrenching architectural changes that Apple has been through before and perhaps again.

In the best of all possible worlds, the good guys and the nice instruction set architectures always win. But in this case the wrong horse won, which made it the right horse in terms of betting after all.

[mercphoto]

But in this case the wrong horse won, which made it the right horse in terms of betting after all.

You missed my point. The x86 architecture is SO UGLY that even Intel tried to get the market away from it, with Itanium. Their goal was a good one. It just so happened that their implementation was poor. Itanium is also an ugly architecture. The market wisely turned down switching from the old but established ugly architecture for a new and unknown ugly architecture. It is quite possible that Intel is unable to design a nice processor architecture. Of course, it doesn't stop them from making money. The book "Accidental Empires" is now floating through my memory...

Adam Smith was wrong. Seldom does the best product win in the market place.

[DavidTO]

Can someone explain to me how the new XBox gets a 3.2ghz PPC processor, but the Mac is stuck at 2.7?

[David_S85]

Transitions are always difficult, but if the end justifies the means, then its good for the company and ultimately the whole market. Apple's on a roll right now, irrespective of their recent stock pricings.

Pentiums have also had their knocks. I can't get much to run on my old PII machine these days either :D . My guess is that Apple won't be using current P4 designs, per se; rather something hybrid oriented, and dual core processor (best of both worlds?).

As I recall, the PPC was a Motorola design and fab, was it not? Apple seeking new biz partners doesn't seem all that odd to me. Might end up being less expensive overall to consumers and Apple than beating a dying horse :deadhorse a few more times (PPC chips). I wrote all this so I could finally use the deadhorse smiley thing.

[DavidTO]

There's also this article (http://www.eweek.com/article2/0,1759,1823273,00.asp) from friday, saying that Apple's going to announce dual core processors.

[ChrisJ]

Can someone explain to me how the new XBox gets a 3.2ghz PPC processor, but the Mac is stuck at 2.7? :rofl:rofl:rofl

The power of the Dark Side is unfathomable.

[mercphoto]

Can someone explain to me how the new XBox gets a 3.2ghz PPC processor, but the Mac is stuck at 2.7?

Because the game machine G5's are not the same as the Apple G5. They are actually different design teams, with different needs and specs. What works for Sony does not work for Apple.

Another person wrote:
As I recall, the PPC was a Motorola design and fab, was it not?
No. Power PC was an IBM architecture, scaled down from their Power architecture for larger, powerful servers. Motorola became a licesee for the architecture (Freescale still is), and the A-I-M alliance was founded. A processor would be designed by a single team of employees from both companies. At one point, as a Motorola employee I had an IBM first-report, and he reported to a Motorolan. For various reasons the alliance fell apart. I was with Motorola at Somerset when the "divorce" happened. Stressful times.

At current moment, all G4's are made by Freescale, and all G5's are made by IBM. The divorce happened during development of the G4, and IBM declined the opportunity to the G4.

[patch29]

Here (http://blog.wired.com/cultofmac/) is another take from a blog at wired.com. We might know shortly.

I could go for a dual core G5. :deal or a super powerbook. :thumb

[Baldy]

When I worked for Steve at NeXT, I was so sick of running on dead-end chips that I called Andy Grove and pitched a NeXTstep (which became OS X) port to Intel architecture.

Intel paid us to do it and helped with the code. It took us 6 months and when we finished, NeXTstep ran better on it than on the sucky Moto chips we were running on.

But nooooo..... Steve wanted to stick with Moto.

Later, Gil Amelio said the reason Apple bought NeXT instead of Be was that NeXTstep ran on Intel chips, which was very important to them, and Be OS didn't. It was the deciding factor.

But noooooo..... Steve wanted to stick with Moto/IBM.

FINALLY! It's about time.

[DavidTO]

Well, it's true! (It's in reverse order, as they post new updates live during the keynote, so it may seem confusing at first)

Under the transformation: Widget/Scripts/Java Just work, Jobs says. Cocoa apps will take a few days to update. Meanwhile, Carbon Apps will take a few weeks. Jobs tells carbon developers to start using Xcode. Over half of developers using Xcode. Next Xcode (version 2.1) will be delivered today. News Xcode generates a single "universal binary" that supports both PowerPC and Intel x86 processors. Available to everybody at registration desk following the keynote.
Mac OS X has been leading secret double life, Jobs proclaims. Every release of Mac OS X has been built for both Intel and PowerPC-based Macs over the last 5 years. Mac OS X is cross-platform by design. Jobs shows all Mac OS X Tiger running on Intel. All features are already compatible with Intel-based processors. It's not done yet, but will be put into the developer hands for finishing.
There is no G5 PowerBook yet. Future products can't be build on IBM PowerPC processors, Jobs says. Intel has better performance and delivers much better performance per watt. Starting next year the first Macs with Intel processors will debut. They'll begin shipping by next WWDC (June). The transition will be mostly complete by 2007 WWDC. It's a two-year transition.
INTEL RUMORS TRUE: Jobs says there have been two major transitions for Mac: 68K to PowerPC and then Mac OS 9 to Mac OS X. Now it's time for third transition to Intel-based Macs. Developers will begin to make the transition now. While users can begin to switch next year. Apple is making the move "because we want to make the best computers for our customers."
Mac OS X "Leopard" is due at the end of 2006

[mercphoto]

I understand your frustration, Baldy, from the Apple side. Rest assured, the Moto/IBM side had just as much frustration with Apple, as Apple had with Moto and IBM. Apple is not a particularly good customer to deal with. There is a reason why IBM and Freescale slowly stopped paying attention to Apple.

[mercphoto]

When I worked for Steve at NeXT, I was so sick of running on dead-end chips

And per that remark, the leading selling 32-bit processor is a Freescale G4. That's correct, more total volume sales than Intel. Hardly dead-end.

There's more to life than personal computers...

[rutt]

Well, a Pentium-M based Powerbook within a year. It's not what Patch was hoping for, but it's good news in the long run for PB fans. Notice that they have Adobe signed up. That was my main question mark. Man, those guys are hard to deal with!

Too bad no Buck Rodgers to make the transition easier. It's going to hurt Apple's computer sales in the short term, I'd think. But what do I know?

[patch29]

Well, a Pentium-M based Powerbook within a year. It's not what Patch was hoping for, but it's good news in the long run for PB fans.

I am glad to hear about future moves which should increase performance are coming for the PB, but now I am unsure about a G5 purchase. I might just keep on waiting. I have waited almost a year so far, what's another year. I would really like a new computer, but with the uncertainty of exactly what is coming, when it will be available, when the software I want will run at full speed on it and how much it will be cost, is a huge ???? Time to sit and wait.

[mercphoto]

Time to sit and wait.

To me, the biggest problem with the switch isn't the switch itself. Its the fact that there is a year before anything is available. Why in the world didn't Steve have something ready to go at the time of this announcement?

I'm a big fan of Macs. I was considering a Mac Mini as part of an entertainment strategy. Run a big 300G FireWire drive off it, store all my CD's in Apple Lossless Compression, run the audio via TOSLink to my receiver. Instant access to all my music at only a small sound degradation over my Arcam CD23 player. Then run the DVI ouptut to a hi-def television, and it also becomes my DVD player.

Would have been sweet. But now I also wonder, why buy a PowerPC-based Mac now?

[patch29]

Would have been sweet. But now I also wonder, why buy a PowerPC-based Mac now?


I cannot see buying one either, unless my PB dies, but it is covered under warranty through this November.

I also am not sure I want to buy the first next generation PB with the first Intel chip, but given that I will have waited almost four years if it comes out in a year I will be due for a new PB and will probably take the plunge. Hopefully they will being to leak news on what new hardware will come out, when, price.......

[ChrisJ]

Why in the world didn't Steve have something ready to go at the time of this announcement?
If the application developers already knew about this ahead of time to port their applications, then the rest of the world would have already known.

My guess is that it's a big announcement to kick-start the application conversion process so that they will be ready in a years time (or so). It sounds like the Operating System itself is good-to-go, and there are developmental Intel based systems waiting to be shipped off.

I had been thinking about getting a G5 for video editing. I'll probably wait on that now. I forsee the price of PowerMac hardware falling faster than normal in the next year.

Maybe this will finally allow Final Cut Pro to run under XP. Seriously, depending on the final architecture, this could mean less work/more money for software developers.

[rutt]

You guys are exhibiting exactly the issue I was interested in. But, you know it had to be done. Better to do it now while all the iPod and other IStuff is hot. If there's still growth there it can cover flatness (or worse) in the computer biz.

Watch ebay for good deals on power gear. People who just bough might want to ditch. And Apple will have to clear the shelves sooner or later.

What do you think the new stuff will be called: InsideBook, InsideMac?

[patch29]

What do you think the new stuff will be called: InsideBook, InsideMac?


I just hope the commercials don't use that Intel noise. :puke

I hope they can deliver it on time. I am definitely in a wait and see mode.

[mercphoto]

I just hope the commercials don't use that Intel noise. :puke

Very doubtful Steve would do that. By bringing in Intel in their ads it dilutes the Apple brand.

[DavidTO]

I just hope the commercials don't use that Intel noise. :puke

I hope they can deliver it on time. I am definitely in a wait and see mode.


BTW: the guy who wrote that little 5 note intel jingle thing is a composer I work with quite a lot. Weird to know the guy that wrote that of all things.

[DavidTO]

Very doubtful Steve would do that. By bringing in Intel in their ads it dilutes the Apple brand.

I don't believe that anyone has any choice about whether or not to use it. They can't advertise Intel products without that thing.

[wxwax]

I don't believe that anyone has any choice about whether or not to use it. They can't advertise Intel products without that thing.
Everything is negotiable. :evil

[DavidTO]

Everything is negotiable. :evil

I'll tell you, if I were Intel there's no way in hell that I would let that kind of PR slip through my fingers. This is the single most significant migration any single computer platform has ever made. What a friggin' coup.

[colourbox]

Everything is negotiable. :evil

The fee for not having an Intel jingle or case badge is probably one of those things Apple will add to the price of a Mac just to maintain its aesthetic purity.

All I know is that Apple is batting 3 for 3 on platform migrations. The survived Apple II to Macintosh, 680x0 to PowerPC, and Classic to OS X. Now they are facing another must-win game...

[DavidTO]

An interesting quote from applematters.com (http://applematters.com/index.php/section/comments/waiting_for_the_macintel_bad_idea/)


It would behoove us to remember that Steve Jobs is a careful speaker when on stage (second hand stories tell us his demeanor is markedly different the rest of the time). An excellent example of his careful attention to subject matter came when Steve was discussing worldwide attendance at the conference. Steve noted that there were a plethora of developers at the conference (3800) but pointedly remarked that there were 38 from China and 11 from India. Why would he mention China and India instead of say, the Isle of Man and Peru? Well it turns out that China and India are huge burgeoning computer markets so it is a prescient idea to motivate the developers representing those countries. Now consider what he said when talking about the Intel chip. Steve noted that while the PowerPC would produce 15 units of performance per watt Intel is promising 70 units of performance down the road. This utterance is notable more what it doesn’t say rather than what it does say. It doesn’t say the Intel chip will be more powerful. It doesn’t indicate that the Intel chip will be faster. What Steve is saying is: given like inputs of energy the Intel will out perform the IBM offering. Which is great for laptops and environmentally friendly but, and this is important, most people won’t care. Many people will happily spend extra money for a chip that consumes 10 times the energy while performing ten percent better. Additionally the metric employed by Steve is also vague. What, exactly, is a unit of performance? So there is no indication that by waiting you’re going to get any performance boost. It is, in fact, plausible that the first generation of Intel powered PowerMacs will (be) outperformed by the machines they replace.

[rutt]

An interesting quote from applematters.com (http://applematters.com/index.php/section/comments/waiting_for_the_macintel_bad_idea/)

David, sorry to be blunt, but you have to look pretty hard right now for a benchmark where the fastest x86 doesn't already beat the fastest PPC. And, as to power not mattering to most people, if those people are the people who wanted but didn't get a G5 Powerbook, those are the ones who don't care about power.

This really isn't subjective. When I was in the processor design business, we paid a lot of attention to the SPEC benchmarks. The people who can perform respectably on this still do. Go look at spec.org. Here's the page I looked at:
http://www.spec.org/cpu2000/results/cpu2000.html. Keep in mind that bigger numbers are better. Apple doesn't play this game, but IBM does. Look at the IBM pServer line. Keep in mind that these are PowerX chips, the big brothers of what Apple uses and quite a bit faster for some things that matter.

[mercphoto]

David, sorry to be blunt, but you have to look pretty hard right now for a benchmark where the fastest x86 doesn't already beat the fastest PPC.
Recently that is true. For the early days of G4 and G5 that was not true. Early G4's tromped the best Intel had. Ditto for early G5's. Not so anymore. Intel caught up while Motorola and IBM were busy paying attention to their embedded market customers. And until Pentium-M came along Intel had a well-deserved reputation for building processors that doubled as frying pans. You think G5's have heat issues?

This really isn't subjective. When I was in the processor design business, we paid a lot of attention to the SPEC benchmarks. The people who can perform respectably on this still do.
Opinions differ. SPEC benchmarks aren't always all that indicative of real-world performance. You can also design and tweak specifically for SPEC. Makes your numbers look good, but won't always make Photoshop run faster.

Power PC and Intel always took rather different approaches to processor design. Intel looked for high clock rates, mostly because it was an easy number to market to customers. But that high frequency means longer and longer pipelines. And x86 code is notorious for frequent branching, which is murder to a long pipeline. Power PC took the approach of shorter pipelines at slower clock rates gets more real work done, with fewer transistors and less power, than the Intel approach. This is the basis for the megahertz myth, and that myth is truth.

To each his own.

[wxwax]

:lurk

[DavidTO]

David, sorry to be blunt, but you have to look pretty hard right now for a benchmark where the fastest x86 doesn't already beat the fastest PPC. And, as to power not mattering to most people, if those people are the people who wanted but didn't get a G5 Powerbook, those are the ones who don't care about power.

This really isn't subjective. When I was in the processor design business, we paid a lot of attention to the SPEC benchmarks. The people who can perform respectably on this still do. Go look at spec.org. Here's the page I looked at:
http://www.spec.org/cpu2000/results/cpu2000.html. Keep in mind that bigger numbers are better. Apple doesn't play this game, but IBM does. Look at the IBM pServer line. Keep in mind that these are PowerX chips, the big brothers of what Apple uses and quite a bit faster for some things that matter.

What I think is interesting is that the computing experience isn't necessarily improved by waiting for the next best thing. Patch29, for instance, may be better served getting his dual G5 now rather than waiting for 2 years for the first high-end MacIntel to come out, which he probably won't buy, since he'll wait for the b revision and for the apps to be updated and stable so he doesn't have to deal with Rosetta.

Over the long haul, this looks to be a good move. But to opt out of buying a dual G5 now, I'm not sure how much sense that makes.

[patch29]

What I think is interesting is that the computing experience isn't necessarily improved by waiting for the next best thing. Patch29, for instance, may be better served getting his dual G5 now rather than waiting for 2 years for the first high-end MacIntel to come out, which he probably won't buy, since he'll wait for the b revision and for the apps to be updated and stable so he doesn't have to deal with Rosetta.

Over the long haul, this looks to be a good move. But to opt out of buying a dual G5 now, I'm not sure how much sense that makes.


I have been waffling back and forth over the last few days. I may still take the plunge on a G5, which with my PB should get me through until rev B on the intel PB comes out. :D

[DavidTO]

David, sorry to be blunt, but you have to look pretty hard right now for a benchmark where the fastest x86 doesn't already beat the fastest PPC. And, as to power not mattering to most people, if those people are the people who wanted but didn't get a G5 Powerbook, those are the ones who don't care about power.

This really isn't subjective. When I was in the processor design business, we paid a lot of attention to the SPEC benchmarks. The people who can perform respectably on this still do. Go look at spec.org. Here's the page I looked at:
http://www.spec.org/cpu2000/results/cpu2000.html. Keep in mind that bigger numbers are better. Apple doesn't play this game, but IBM does. Look at the IBM pServer line. Keep in mind that these are PowerX chips, the big brothers of what Apple uses and quite a bit faster for some things that matter.

Also, keep in mind that in real world terms, the current G5s could perform better than the Intel's, especially if you end up having to run Rosetta. Rosetta does not support Altivec, so you will in effect be running a G3 processor. In real terms, the benefits of the Intel move may not be immediately apparent to the average user. In theory Rosetta can run at 80% speed, but that's just theory. Real world numbers will most likely be slower.

[DavidTO]

patch29,

Here's more rationale for you to buy now:

macintouch.com (http://macintouch.com/#commentary.2005.06.09)

[wxwax]

G5 prices be affected in the short term?

[patch29]

patch29,

Here's more rationale for you to buy now:

macintouch.com (http://macintouch.com/#commentary.2005.06.09)

That is close to what I have come up with. All my software works, the performance will be very good and a huge jump up from a Ti PB 1ghz 1gb ram to G5 DP 2.7 8gb ram. It should last me through the hardest part of the transition and if it gets off to a rocky start my system will still be chugging along. :thumb

Where did I leave that credit card? :scratch :deal

[patch29]

G5 prices be affected in the short term?


There has not been any affect so far. I would think if sales continue and orders come in they will not drop them until they are just about to be replaced which could be 1-2 years.

[DavidTO]

That is close to what I have come up with. All my software works, the performance will be very good and a huge jump up from a Ti PB 1ghz 1gb ram to G5 DP 2.7 8gb ram. It should last me through the hardest part of the transition and if it gets off to a rocky start my system will still be chugging along. :thumb

Where did I leave that credit card? :scratch :deal


I've got the exact same configuration in my Pbook, and the dual 2.0 G5. You are going to be very, very happy with the performance.

[patch29]

I've got the exact same configuration in my Pbook, and the dual 2.0 G5. You are going to be very, very happy with the performance.

I have a friend with a dual 2.5 and I have run a couple tests on it and I know I need to do it. It is just a lot of money and I am looking at possibly adding a 4800, a lot of lighting gear, a 90mm TSE lens and a few other goodies, so it is getting expensive very quickly, at least I won't be short on write offs this year. :uhoh

[rutt]

I have a friend with a dual 2.5 and I have run a couple tests on it and I know I need to do it. It is just a lot of money and I am looking at possibly adding a 4800, a lot of lighting gear, a 90mm TSE lens and a few other goodies, so it is getting expensive very quickly, at least I won't be short on write offs this year. :uhoh

I've been thinking this through, and I think it's the right thing to do. If you were looking for a new PB, that would be a different thing. But the high end x86 machines are presumably a couple of years away, which is a long time to wait. In the meantime, for photoshop, the dual 2.5 is as fast as anything with Intel chips except Xeon which is pretty expensive (though I've seen some low prices recently.) So I think this thing will be a workhorse for at least a couploe of years.

Once Apple has high end Intel products, these are going to be very tempting. They need to make their first high end intel machines be real performance monsters and it's only a matter of shopping Intel's road map to see that this will be more than possible. What about 2X quad core Xeons at about 8GHz?

As to Powerbook, right now my Pentium-M linux notebook is more than 2x my G5 Powermac (1 processor) for almost every test I can come up with. So this is also going to be a very nice computer, and I think it will be among the first things Apple releases (though the first ones will be 32 bits only.) After all, this was the strongest motivation for the switch.

BTW, definitely shop ebay, etc, for your new Powermac. The supply is rising faster than the demand.

[ian408]

All of this banter has been very interesting.

Jobs is shrewd. I don't think he'd announce a change like this unless
the time to market were a year or less.

Why do I think this? By announcing a switch to Intel and implying it's for
power consumption as much as performance, the bottom will fall out of PB
sales for a time as it will for desktops although maybe less so. I know that
one of the factors that keep me from buying a PB is the lag in processor
technology between the desktop and PB.

Apple marketing will have a tough time promoting existing technology
during the coming transition. I would think that pricing will be more
favorable in 6 months.

Ian

[rutt]

Jobs is shrewd. I don't think he'd announce a change like this unless
the time to market were a year or less.
Among the shrewdest. His only problem is massive ego. But that just levels the playing field...

.. the bottom will fall out of PB
sales for a time
My guess is that this has already happened to some extent, so there wasn' that much to lose. Think about the advice I just gave Patch. Powermacs still make sense because their performace is OK vs Intel bases machines, especially on a price/performace basis. Give them a 2 year lifetime, say. It's enough. The iMacs make sense because the package is so nicely differentiated and the price is good and they aren't really performance machines. Macmini? I don't know vs a linux machine in a shuttle package, but for Mac addicts? PB, well, I've been saying for 1/2 a year that it's time to wait or try to figure out if you can use a windows or linux notebook.

[ian408]

I think your average iMac user will probably not buy a new computer for a
really long time--until the hardware chokes and pukes.

On the other hand, Shuffle/Ipod customers probably consider those devices
throwaway and will buy when the next new thing when it comes along.

DP G5 users are probably careful but like Patch, will buy based on need.
Knowing that it will be a two to three year investment and that a lot
happens between initial purchase and end of life, will weigh their software
cost accordingly.

I would agree that PB sales may have slowed. I think some are waiting for a
G5 based machine and that will probably not happen unless you have the
room for a small cooling tower :D

Ian