Slashdot stories can be listened to in audio form via an RSS feed, as read by our own robotic overlord.

 



Forgot your password?
typodupeerror
Technology (Apple) Businesses Apple Technology

Intel Inside For Apple? 239

Posted by pudge
from the slow-week dept.
iomud writes "Bear Stearns analyst Andrew Neff predicts that there's a better than 80 percent chance Apple will make the jump to Intel in two to four years. As the relationship with Motorola seems to be weaning the question may be what chip would you like to see in next-generation Macs and why?" It seems important to note that Bear Stearns owns shares of Intel and Dell, and has a banking relationship with Dell and HP. Oh, and even if it didn't, that I can't see any reason why anyone should care what Andrew Neff says. But that doesn't mean it can't be fun to talk about!
This discussion has been archived. No new comments can be posted.

Intel Inside For Apple?

Comments Filter:
  • "Neff, for instance, predicted Apple, which uses chips from Motorola and IBM that currently top out at 1GHz, will switch to Intel, whose chips run at 2.5GHz, to get a performance boost and gain more customers. There's a better than 80 percent chance Apple will make the jump in two to four years, he said." This seems to imply that the 2.5 GHz P4 is 2.5 times as fast as the 1 GHz G4... Which is a joke. However, a lot of people (primarily the ones buying their PCs at Walmart) are great believers in the MHz Myth and will compare the two chips based just on clock speed. This indeed might make more gain in terms of customers for Apple, but at what cost? Chips that run hotter and process fewer instructions simultaneously? How about instead of advertising chips in terms of clock speeds, start marketing them in terms of calculations per second (start comparing gigaflops... in which case, last I checked, G4s were way ahead of Pentiums). -T
    • How about instead of advertising chips in terms of clock speeds, start marketing them in terms of calculations per second

      You mean, cycles per second WHICH is Hz. Thus a Pentium IV at 2.5 MHz is 2.5million cycles per second.

      Enough said
    • The "MHz Myth" is a myth, or more accurately RDF. MHz do matter but they are not everything. Historically the PowerPC has held up extremely well against Intel. Some programs really do excel on the PowerPC but in general you get about a 20-30% increase when comparing PowerPC and Intel of the same clockrate. Assuming properly compiled and equivalently optimized programs, no Apple PR games like using old 486 optimized code on a Pentium (ByteMark), G4 vs. Pentium 4 comparisons where the Mac code uses Altivec and the PC code does not use SSE2, etc.

      If someone wants to argue that there is practically no difference between a 1.0 GHz G4 and a 1.4/1.6 GHz Pentium 4 I would readily accept that. You need a benchmark program or a good stopwatch to tell the difference. However with Pentium 4's up to 2.5 GHz (and 2.0/2.2 GHz being pretty inexpensive) you will find that raw brute force MHzs does matter. It may not be the 2.5:1 that the non-technical might assume, but it is noticable.

      Comparing CPUs in terms of operations? Well that's what SPEC is all about. However Apple does not like SPEC since it is not RDF friendly and contradicts the arguement that MHzs don't matter.
    • It's true that you can't compare chips directly, but try an actual benchmark (ie, not a photoshop filter commissioned by apple) and you'll see that the 2.5 GHz P4 beats the 1 GHz G4 pretty easily. Try cpuscorecard.com, for instance, which says a 1 GHz G4 is a little worse than the 2 GHz P4.
      • The ultimate smackdown: Mac versus PC [usatoday.com]

        Dual 1GHz G4 versus 2.2-GHz Sony Vaio RX690G Digital Studio.

        "Rather than argue the point, I decided to conduct my own comparison. Apple even provided the actual Photoshop picture used, a full-color photo over 44 megabytes in size, depicting seven bike riders with colorful bikes and helmets. This is the sort of file that graphic artists have to manipulate on a daily basis.

        "Apple also sent me a copy of its actual test protocol, including a Photoshop Actions file, a set of scripts that automated the various rendering functions. They also provided a high-end twin-processor desktop Power Macintosh on which to run the tests.

        "The Mac was upgraded to the latest version of Mac OS X, 10.1.5. The Sony had Windows XP. I installed the standard retail versions of Adobe Photoshop 7 on both computers."

        "Running the tests proved exceedingly simple because Photoshop displays the actual timing of a rendering process rounded off to tenths of a second. Per Apple's directions, I conducted each test four times to deliver the most accurate results.

        "Like all Adobe applications, Photoshop is a bit slow to launch. It took 15 seconds on both computers to get ready for the main event.

        "In the nine test runs, the Mac came out on top five times, besting the Sony by up to 8.1 seconds. Where the PC emerged victorious, the margin was usually less than half a second.

        "In all, the Mac took a total of 35.5 seconds to complete the nine rendering steps. The PC took 50.1 seconds, making it 41% slower according to my calculator's reckoning.

        ...

        "The upshot of all this, however, is that, when someone tells you a Windows box is always faster than the Mac, point them to this article and tell them it isn't necessarily so."

        Of course the PC beat the Mac in a game of Quake ;)


        • It's true that Photoshop has filters optimized for the Mac, and the benchmarks for those are somewhat irrepresentative of general performance. Most programs do not have this kind of parallelism available, and even fewer are actually optimized to use the processor's vector capabilities. (A better benchmark would compile the same C program using the vendor's compiler on both platforms and measure how the two stacked up. Or at least allow both vendors a shot at optimizing the filters in question...)

          But the main problem with this test is that he's testing a dual processor G4 against a single processor Pentium in a multi-threaded app doing highly parallelizable work!! How can we make sense of those results?

          • But the main problem with this test is that he's testing a dual processor G4 against a single processor Pentium in a multi-threaded app doing highly parallelizable work!! How can we make sense of those results?

            easy, it is 2 off the shelf systems, you plug it in and see that one system performs better then a diffrent system. (last i checked, there wern't any dual P4 systems avalible.) this test shows that apples high end system beet out sony's system (i'm asuming it's their high end system but i didn't read the artical)

            the test i'd like to see is apples high end system up against a high end athlon system 1, 2, 4 CPU's it doesn't matter, the athlon will smoke the apple in perfomance, but the apple will smoke the athlon in usability.
          • It's true that Photoshop has filters optimized for the Mac

            Actually the filters are optimized for Intel processors too. Intel gave Adobe the assembly code needed (for MMX at the time, and I'd imagine newer code since). IIRC only two or three PS filters are optimized for AltiVec. I think Photoshop is used a lot because it's the same code base on both platforms.

            And this isn't all about filters. There was a lot of transforming and compositing going on.

            But I agree that some things run better on one platform or another, and it might have more to do with the OS than the CPU type, and how well a program has been written for that OS (i.e. MS Word).

            But the main problem with this test is that he's testing a dual processor G4 against a single processor Pentium in a multi-threaded app doing highly parallelizable work!! How can we make sense of those results?

            But the single Pentium was more than twice the clock speed, and we know that a dual 1 GHz computer is not 2 GHz, right? :) Photoshop is kind of buggy on dual G4s... its been known to lock up a lot, and some people remove the MultiProcessor Support Extension. In some tests single CPU G4s outperform the duals.

            And as the tests showed, the PC did better at Quake, and other tests on the 'Net have showed PCs outperforming G4s in Adobe After Effects tests (AE runs like a dog in OS X)

            The bottom line is that clock speed isn't everything.

          • But the main problem with this test is that he's testing a dual processor G4 against a single processor Pentium in a multi-threaded app doing highly parallelizable work!! How can we make sense of those results?

            Part II

            Here's something interesting: [linuxworld.com]

            Cheaper & faster, too
            Several readers, including an Apple sales rep, sent me references to a set of Xserve benchmarks [apple.com] on the Apple site. All of these show the Xserve beating competitive models from other companies, including IBM, Sun, and Dell.

            What makes two of these results particularly interesting is that they show the value of optimizing software to take advantage of the hardware, reversing an effect I think of as "regression to the dumb" to achieve impressive results.

            "Regression to the dumb" reflects, I think, the marketing tendency to focus on simple things that are easy to communicate in a volume market and elevate these simplifications to the level of de-facto standards. Engineers then have to accommodate these standards in product or process design.

            The "megahurtz" wars, long a sore point for both Mac and Sun users, seem to illustrate this perfectly. Each new generation of x86 CPUs does less per cycle than the one before, but it drives the claimed megahertz number up because that's the number that moves product. Along the way, some very good technologies have been abandoned, and software developers have been taught to avoid making their code dependent on chip-specific features that could easily go away with the next iteration.

            What happens if you look carefully at the technical advantages you've got and optimize your code and hardware accordingly instead of just going with industry-averaging practices?

            1. Apple's PowerPC has an underappreciated facility marketed as the "velocity engine." This is actually a short-array processor with powerful features such as hardware FFT, but, like SPARC's VIS/SIMD, it's more honored in the breach than the observance.

              In this case, Apple's Advanced Computation Group, working with Genentech, modified an application widely used in genetics and related research to make maximum use of the facility. As a result, the Blast benchmark, which searches a genetics database for matches, shows the dual 1-GHz Xserve beating an IBM x330 with dual 1.4-GHz P3 CPUs by factors ranging from 5.8 to 21 (and a Sun V100 by up to 52 times) depending on the length and precision of the matches.

            2. Internally, the Xserve has DDR (double data rate) memory feeding a 4-gigabytes-per-second data path to the CPU cache along with four ATA controllers -- one for each disk -- that operate as one. Using Bonnie [textuality.com] to compare I/O to a Dell 1650 with dual 1.4-GHz P3 CPUs; SDR (single data rate) memory; and, a single Ultra160 RAID card with 128MB of buffer, Apple finds that the Xserve can be more than twice as fast as the Dell.

              Technically, I believe that there are two factors at work here: the Xserve has faster memory and a cleaner data path to the CPU, and Apple's four-way ATA design is both faster and cheaper than the single-path RAID card.

            In both cases, better technology used in smarter ways wins. As in, duh? But managerially what they've done here is pretty cool because they're standing up for excellence instead of collapsing the technical tent and going off in search of volume.

      • 1) All valid benchmarks show that PCs are faster than Macs.
        2) All benchmarks that show something else are thus invalid.
    • It would be really nice if Apple, in fact, showed some real, industry-standard benchmark results to support their performance claims, but they don't.

      When others have looked at the G4 performance on a standard benchmark suite like SPEC (e.g., here [heise.de]), a 1GHz G4 is not significantly faster than a 1GHz Pentium III.

  • False Information (Score:2, Redundant)

    by Sandman1971 (516283)
    It's hard to take this article seriously when it attempts to spread false information.

    Neff, for instance, predicted Apple, which uses chips from Motorola and IBM that currently top out at 1GHz, will switch to Intel, whose chips run at 2.5GHz, to get a performance boost and gain more customers. There's a better than 80 percent chance Apple will make the jump in two to four years, he said.

    Everyone knows you can't compare speeds of Intel and Motorolla chips, as they do not equate to the same thing. I lost all respect and believability for the article after reading that piece of rubbish.
    • Everyone knows you can't compare speeds of Intel and Motorolla chips, as they do not equate to the same thing. I lost all respect and believability for the article after reading that piece of rubbish.

      Of course you have blown your credibility with the above as well.

      MHz can't be used as a precise measurement but it can not be completely disregarded. Especially when the ration is over 2.5:1. Is a 1.4GHz Pentium 4 faster than a 1.0GHz G4, for all practical measurements probably not. A 2.5 GHz Pentium 4, yes, raw brute force can overcome elegance and efficiency.
  • Yeah, right. (Score:1, Redundant)

    by Henry V .009 (518000)
    Neff, for instance, predicted Apple, which uses chips from Motorola and IBM that currently top out at 1GHz, will switch to Intel, whose chips run at 2.5GHz, to get a performance boost and gain more customers. There's a better than 80 percent chance Apple will make the jump in two to four years, he said.
    I'd like to make a brief, stunningly persuasive, riposte to his argument:
    Yeah, right.
  • Nope (Score:1, Insightful)

    by Anonymous Coward
    Apple likes to build cool stuff. Noise is not cool. I don't think we'll see Intel based Apple machines any time soon unless there are drastic strategic changes at Intel.
  • What's the point? (Score:1, Insightful)

    by Anonymous Coward
    Why think about this now? Apple just moved to a totally new operating system in which only 20% of their user base has switched.

    Additionally, the size of the Mac user base has steadily eroded but there are marked decreases around both the introduction of System 7 and the PowerPC chip. To switch now would be suicide! Apple may indeed want a different processor, but doing so would probably mean that applications would have to be rewritten and we all know how long it took to get Photoshop out the door and many people are still waiting for Quark.

    If they do switch, then good for them. History would suggest they should wait a while before undertaking such an effort and in the meantime this is just intellectual masturbation, IMHO of course.

    Unfortunately this gentleman raises no good points other than the disparity between the processor speeds. Don't get me wrong, I am not someone who has been blinded by the MHz Myth as brought to you by the Reality Distortion Field, but his arguments are nonexistent. The fact that he has predicted a few other industry actions is anecdotal at best and irrelevant at worst.

    Short version: Take this guy worth a grain of salt. Wait a year or two and see what the processor landscape looks like.
    • by BitGeek (19506) on Monday August 05, 2002 @06:15PM (#4015183) Homepage
      Additionally, the size of the Mac user base has steadily eroded

      I don't think you can say this. I'm aware of no information that expresses the size of the mac user base.

      you often see the "%5 of the market" figure, but that is actually %5 of NEW PC SALES, (so it ignores the fact that People turn their PCs over every 18 moths, but macs are performance competitive a lot longer) oh, and these numbers also ignore most mac sales. So even saying "%5 of new sales" is a lie-- they count Dell, Ingram Micro and CompUSA. They ignore the Apple store, the Apple stores, and the hundreds or thousands of independent apple dealers around the world.

      Put a better way, Apple has %5 of the Intel PC market- - because that's the market they count-- and of those people, %5 of the pcs they sell are actually apples!

      The total addressable market-- that is, Macs out there in active use-- is much larger, probably %20.

      Last time I had any reliable numbers, it was %30, but that was because they were the only company selling CDROM drives for computers and so you could look at the number of those sold and know how much market share apple had... so that would have been the early 90s.

      I'm not saying I know what the TAM for Macs is, I'm just saying I've never seen any reliable figures, and the %5 one is clearly unreliable. ( But makes for good copy for those with "Apple is dying" stick who want to beat that dead horse.)

      • Actually, the "Apple is dying" crowd usually say 2%-3%. The 5% figure has often been displayed prominently in Apple's own ads. If that figure ignored Apple's in-store and on-line sales, don't you think Apple would have commissioned another study by now, or demanded a correction from the companies doing these surveys, rather than run adds on their website saying, "now if we can just convince 1 out of every 19 PC users to switch to a Mac, we would double our market share!"

        • I've only ever seen apple use the %5 figure when they were JOKING.

          "Now to get the other %95" is a clear joke to me, though most people seem not to get it.

          They are playing on the perception that they have %5 of the market.

          Anyway, the fact is that the %5 figure comes from a fundamentally flawed study of the marketplace, and I have not seen any better research.

  • Not clawhammer (Score:4, Insightful)

    by Perdo (151843) on Monday August 05, 2002 @06:00PM (#4015092) Homepage Journal
    Sledgehammer. Opteron. Whatever.

    Not Itanic.

    Not Pentium 4

    Not C3 (heh, I just benched a C3 800. It performed about as well as a 266 PII except with the P4's weird imbalanced interger performance. the numbers looked about like a P4@500mhz)

    Stick a few Opterons in an Apple and you take Apple back to the good old days where their hardware actually outperformed the x86 boxes and was still somewhat unique.

    Let Apple shine again... not just on the outside, but on the inside too!

  • Switch and die (Score:4, Insightful)

    by jpt.d (444929) <abfall@UUUrogers.com minus threevowels> on Monday August 05, 2002 @06:00PM (#4015097)
    if ([apple switchTo intel])
    [apple killSelf];
    • I think your Objective C is a bit off. Try:

      if ([apple switchTo:"@Intel"])
      [apple dealloc];
    • *if* Apple switched to Intel chips, would this somehow invoke some sort of pressure from MS on Intel? youu figure if OS X ever ran on straight up PCs (doubt it) then they would be going head to head with Bill Gates and i see him fighting back. if they used a modified intel chip (doubt it), then i wonder if it would matter if M$ has their foot int he door of the plant that makes Apple's chips. also, why bother? Intel chips are hot as hell and use tons of power. these are the days of power usage concerns, and Apple's dislike for fans and noise. if anything they should be pitching the power usage of the LCD iMac vs some P4 with 19" CRT device. granted on one user's house it isn't a big deal but when you have clusters and clusters of them in schools and offices it adds up.
  • by BitGeek (19506) on Monday August 05, 2002 @06:01PM (#4015101) Homepage

    We all know that PowerPC chips get far more done in a given clock than x86 chips.

    This was the great promise of the PowerPC, actually. By going to a superscalar Risc architecture, IBM and Motorola spent the effort to get a chip that really did more per clock.

    The clock rate, however, is less of an engineering issue than a process issue. Intel has processes that increase their clock rate rather fast-- and so rather than re-engineering their processors (and paying the backwards compatibility penalty that apple paid when they switched from 68k to PPC) they have simply increased the clock rate and integrated more on chip cache, etc.

    The thing is, this means that the PPC was at a very significant competitive advantage-- its really hard to beat architecture engineering, which the PPC has in spades, but pentiums lack. Design is hard. Process is easy. So, the Processes that Intel was using should have migrated to Motorola and IBM, and we should be seeing PowerPCs that run at 2GHz and leave no question as to the fact that the powerpc is much much faster.

    So, the real question to my mind is-- why hasn't the process side of the house for PowerPCs kept up with intel? Certainly motorola and IBM have the know how, and they have the motivation-- competition with each other for the sizable sales to Apple, and the possibly even larger embedded and workstation markets.

    I can think of two possibilities:
    1) The increased complexity of a super scalar architecture on the order of the PPC makes timing more problematic and while process is there for higher speeds, the synconization of the clocks hitting all the subcomponents of hte processor at the same time is an issue. At these levels, the speed of light is a real factor when one signal goes a little further than the other, they arrive at the same place at different times due to the relative slowness it takes for the signal to go down the longer path.

    2) Conflict. Motorola created Altivec and apple jumped all over it, and I don't believe IBM has a license to Altivec, giving motorola a bit of a monopoly. This combined with apple embracing altivec so much means that Motorola may not have sufficient incentive to grow the speeds. Plus, since the PowerPC has not had the widespred platform support that was expected-- NT for PPC has gone away, other Unix box makers aren't using it extensively, the market is smaller than was originally intended.

    This creates quite a problem for apple. As long as they suffer from the perception- despite the reality-- that their processors are slower because people think MHz = speed-- they are going to have trouble not being seen as more expensive. Hell, even people who post here make this mistake.

    So, I think Apple is planning something big. But it won't be a switch to x86, certainly as we know it.

    I can imagine a couple possibilities:
    1) Apple teams with AMD and brings the PPC instruction set to a future AMD processor that can handle it and the x86 instructions simultaneously. Gets AMD's process speeds, along with PPC compatibility running at native speeds (rather than emulated.) The downside is that IBM would have to agree to this, and its not clear what IBM's upside is-- unless IBM is part of the alliance and gets a competitive advantage to using this technology in its products (maybe low end power workstations)-- but still Motorola which controls altivec would have to be involved.

    2) A new AIM partnership, this time its the AAIM partnership, all four companies collaborate on a new chip that will run OS X and Windows, IBM and Moto make PCs that dual boot, AMD gets Altivec and Power4 Multichip module technology, and IBM and Moto get AMD process technology, and IBM, Moto fab the chips for AMD. This gives IBM a weapon against windows, namely OSX, gives AMD the backing of two big competitors- IBM and Moto, along with a new customer, gives Moto a new jumpstart into the box making business that it gave up when Apple stopped subsidizing the clones industry.

    3) The Death By Numbers Approach -- Apple goes to IBM and gets the four chip Power technology and migrates there from PowerPC, greatly increasing the volumes of these chips for IBM which is only currently using them in their servers and workstations. This drives down the costs, apple doesn't have to rewrite software (like quicktime) that was never part of the NeXT OS, and at the same time can emphatically claim the "fastest PCs in the world" title it now holds but nobody recognizes. Oh, and they sell them with 2 to 4 processor units per box.

    4) Death By Numbers part 2-- apple starts shipping quad and 8 way PowerPCs running at moderate speeds, 1-2GHz using Motorola (or IBM) chips, and being competitive on price because the powerpc costs them so much less per cpu than Intel CPUs. Thus people will instinctively know that 8 1GHz CPUs are going to get a lot more done than one 3GHz intel cpu.

    5) The Second Rebel Alliance-- Apple, AMD and Nvidia team up on an x86 processor that uses NVidea and AMD Hyper IO (or is it rapid io?) technology, and apple does go the x86 way..

    The thing is, 5 seems least likely to me. apple has just migrated accross platforms for the second time-- the first was 68k to ppc, and the second is classic Mac to OS X. Applications have to be re-written.

    Are they really going to ask their developers to re-write their apps yet again, in only a few years? I really doubt it.

    So, I think there is a new processor architecture or solution coming-- I'm sure apple recognizes that the PPC has not given it the marketability it needs.

    But I think that solution will be PPC compatible natively.

    • What about spec benchmarks [queru.com]? I know ppc is a fast chip but I just dont think it's as fast as it used to be. Along with moto's reluctance to put out literally anything lately makes me very concerned about the future of the chip. There's only so many ways you can repackage year old technology, apple.. aside from the xserve is struggling in the hardware department. The price goes up but the performance doesn't. That being said, I own a 933g4 powermac, I bought it about three months ago, it's system bus is slower than the pc system I built last year, just a bit discouraging.
    • by danielwright (114541) on Monday August 05, 2002 @08:41PM (#4015745)
      You're oversimplifying things a little too much when you say that if a PowerPC chip were made with the same process that Intel uses for it's new P4s, it would have the same clock rate.

      Modern CPUs are all pipelined, so they divide each instruction into several pieces - say Instruction Fectch, Instruction Decode, Execute, Load/Store, Write Back for example. Then, they interleave the execution of the different stages, so while one instruction is being decode, the next is already being fetched.

      At a very rough approximation (it's much more complicated than this), the clock rate has to be low enough that the largest of the pipeline stages can execute in one clock tick, so if tou divide up the execution into more, smaller stages, you can raise the clock rate higher. However, there's a lot of complex machinery to avoid "hazards" where instructions depend on each other, so they have to stall some of the instructions, and this gets more complicated and slower with a longer pipeline. (This would be a gross simplification 10 years ago, and today's CPUs are much more complicated, but it gets the main point across).

      The designers of the current PowerPC implementations chose fairly short pipelines (I'm not sure of the number of stages, but I think it's around 5), while Intel uses 20 stages for the P4. That means that the P4 can run at a higher clock rate, but get less done per cycle because more of the instructions are stalled.

      So, my point is, at least IBM has CPU processes at the same level as Intel's, if not better - it's due to the fundamental design of the chip that the GHz number is lower, which makes the GHz a very uninteresting measure - hence the "MHz Myth".

      Also, PowerPC is an instruction set, like IA32 or IA64, it's not a chip architecture. IBM and Motorola currently make chips that implement the PowerPC instruction set (and IBM's chip, the Power4, is currently the fastest chip available, BTW).

      Just to add to the list of totally unfounded predictions, here's mine:

      IBM released the Power4 a few months ago, as the fastest chip on the market. They want to use it for every server platform they make (AIX boxes, mainframes and AS/400 boxes). It's designed for servers, and that shows - you need something like 1 ton of force to attach it to the motherboard, and a pretty impressive cooling system as well. This makes it unsuitable for small desktop machines like the imac, and for laptops. Also, it doesn't support Altivec. I figure, they'll work out some licensing agreement so they can make a special, slightly slower version for Apple that does support Altivec.

      The merits of this: they could use basically the same CPU design and processes (which are very, very good), and now software changes.

      I don't think Apple can change to Intel chips because that would require new versions of all the software. They've just asked all their customers to replace old OS9 software with OS X software. If they came back in 2 years and said everyone should replace all their software again, their customers would start to get rather irritated by it...
      • Also, PowerPC is an instruction set, like IA32 or IA64, it's not a chip architecture. IBM and Motorola currently make chips that implement the PowerPC instruction set (and IBM's chip, the Power4, is currently the fastest chip available, BTW).

        No. Power4 uses a slightly different instruction set. Running PPC code on a Power3 or Power4 chip would require recompile or emulation. It's a lot closer than x86 or something, so emulation might not be the worst idea, but it's definitely different.
    • We may all "know" that, but it seems to be a myth. At least on SPEC benchmarks [heise.de], a 1GHz G4 PPC doesn't do a whole lot better than a 1GHz Pentium III. The SPEC benchmarks are a pretty good mix of real-world code. What Heise got on them is probably what you and I can expect when we compile our programs. One might also note that, despite Apple's constant claims about how powerful the G4 is, they have never submitted a SPEC benchmark result for the chip themselves.

      I think the PPC is a dead end for Apple. Lack of a 64bit migration path is a problem. Intel's Itanium doesn't need to fear comparison architecture-wise with PPC either. But the mainstream will go to 64bit AMD and Pentium. That's perhaps where Apple should go as well.

      • IBM is already on the second generation of 64 bit PowerPCs (POWER4). They can run either 32 or 64 bit code. Basically, PowerPC already made the transition that x86 is just beginning now.

      • What Heise got on them is probably what you and I can expect when we compile our programs.

        Depends on the compilers used.

        One might also note that, despite Apple's constant claims about how powerful the G4 is, they have never submitted a SPEC benchmark result for the chip themselves.

        One might also note that for years SPEC simply didn't run on Macs (not even running any *NIX). It may still not work too well on OS X (or Linux PPC).

        I think the PPC is a dead end for Apple. Lack of a 64bit migration path is a problem.

        PPC has been designed from the start to have both 32 and 64 bit implementations. And IBM does consider Power4 to be a PPC chip.

        Intel's Itanium doesn't need to fear comparison architecture-wise with PPC either.

        If you don't mind only being able to use one and only one compiler. At least if you want any speed.


      • I don't consider the "SPEC benchmarks" to be a very good citation-- there are a variety of benchmarks in SPEC, and they certainly don't reflect the instruction mix of modern applications.

        For instance, penitums are really good at doing integer calculations but very poor at floating point, yet almost all applications that are CPU INTENSIVE use floating point. Yet Spec gives integer a much higher rating, and generally ignores floating point optimizations that are used in real world situations.

    • by GORDOOM (149962)
      The really frustrating part of this whole mess is that IBM has been able to get G4s up to some mightily impressive clock speeds - not as high as the P4, perhaps, but certainly higher than any of Motorola's G4s - but they're not allowed to sell them. Why? Because Motorola doesn't want IBM selling G4s faster than it can make them, because that would make Motorola look bad. And since Motorola owns the Altivec routines, IBM has no recourse. And so, now Apple is caught in the middle of this mess, stuck with slow G4s!

      But Motorola has been having problems of late, and may be willing to sell off parts of the semiconductor division... if Apple could buy the code for the Altivec routines from Motorola, and then licence that code to IBM, but without the restrictions on processor speed... I wonder how fast IBM could get the G4 running then... :)
    • The thing is, 5 seems least likely to me. apple has just migrated accross platforms for the second time-- the first was 68k to ppc, and the second is classic Mac to OS X. Applications have to be re-written.

      Are they really going to ask their developers to re-write their apps yet again, in only a few years? I really doubt it.


      i believe that you have hit the nail on the head with this point. architecture shifts are huge and take a long time to complete. apple has pulled it off once, 68k -> ppc was quite well done. os9 -> os x has just started, but is going quite cleanly, given it's drastic nature.

      i think your 2nd option is the most likely, but nvidia needs to be in there somehow. (AMAIN??? ;)

      what i think would be quite interesting, is if apple took an even bigger roll in developing chips. they have alot of knowledge in-house, and partnering even closer to people like amd and ibm might be a good idea.
      so servers/workstations get those giant power4 chips, and portables/consumer machines get the g(4||5).

      a remaining question is what is apple's 64-bit strategy?
    • Excellent, most excellent presumptions.

      It should be noted that, of any personal computer, only Apple can even consider such moves without significantly affecting (adversely) the potency of their computers. No other mobo spec maker can, or has, dramatically changed their systems in the way that Apple does.

      I presume the same, that is, that Apple is seriously considering a processor change. It may be for performance, but the decision will also be for a cost advantage. ANYTHING to reduce the cost of a Macintosh yet provide the same performance and convenience is a Good Thing for Mac sales.
  • My wishful-thinking-cap is still firmly pointing at Apple ditching Motorola and going to IBM for their processors.

    a POWER4-Lite would be vaguely feasible (eg, pair of G3 cores + SMP logic + Altivec execution hardware + 1MB of L2 cache) on .13, and I'd reckon it would be rather rapid :)

    Of course, the chances of that happening are something like my chances of winning the lottery, which incidentally is also the only way in hell I could afford a PowerMac equipped to my liking :p
  • It is feasable for Apple to put a Pentium on it's motherboards as a co-processor. The extra prossessor could get used by apps that need another floating point unit. Normall, non processor-intensive apps could just ignore it.

    It would be a stupid hack, but woulden't require any recompiles for curent apps and gould get rid of the 'MHZ Myth' once and for all.

    Of course this would be non-elegent, and mostly for marketing reasons.
    • It would be extremely non-elegant, and if the 68k / PPC hackjobs for the Amiga are anything to go by, it would be hellishly slow too..

      Maintaining cache coherency between two processors that are opposite-endian... eek *shudder*
      it was bad enough with the 300+ _Micro_second context switches on Amiga's with PPC accelerators
  • I see Apple making Mac OS XI for x86 but only allowing it to work on special Apple motherboards. Apple won't hype the switch that much. They will instead sell some sort of VMWare-like or dual-boot stuff and market the x86 Macs as being able to run Windows at full speed.

    Then someone will hack Mac OS XI to work on any motherboard, or some company will reverse engineer the special Apple motherboards and make their own Mac compatible motherboards, and Apple will call out the lawyers.

    Tim
    • Re:The future (Score:3, Insightful)

      by feldsteins (313201)
      I see Apple making Mac OS XI for x86 but only allowing it to work on special Apple motherboards.

      I think that is exactly right.

      Apple won't hype the switch that much. They will instead sell some sort of VMWare-like or dual-boot stuff and market the x86 Macs as being able to run Windows at full speed.

      Can't see it. What I do see is that Apple will make the switch when a next-gen Intel or AMD processor comes out - and they will wait for it for two reasons. 1. Presumably one of them will find a way to make their stuff a little smaller and cooler. Apple likes things like TiBooks and fanless iMacs. Can't develop shit like that with brick-sized P4 modules can you? No. 2. Apple won't want to pull a "New Coke" on their market. Mac users are loyal to their brand and to their processors. They won't like seeing a switch to a part that has been touted as inferior for so long. This effect will be lessened when a next-gen part comes out which doesn't have quite the history of being bashed by Apple as the current one's do.

      Then someone will hack Mac OS XI to work on any motherboard, or some company will reverse engineer the special Apple motherboards and make their own Mac compatible motherboards, and Apple will call out the lawyers.

      Apple would never, ever make such a switch unless they were supremely sure that this couldn't happen. If the ability to sell proprietary hardware for the OS went bye-bye then so would Apple itself and they are fully aware of this. It's not just a dinosaur clinging to the old ways...it really is at the core of Apple being able to innovate the way they do. They have to control the OS and hardware of the platform to do what they do. That is the only reason why Dell or Microsoft can't be an Apple. it's not because Apple is "cooler" or even "smarter." It's because they control the entire platform.

      Hell, if I worked at Apple I would want to make damned sure that those crown jewels never got lost. I'd rather run the boxes with hampsters in plastic wheels than risk that.
  • I love how the article states that Neff says that the HP Compaq deal was a bad idea:

    HP, meanwhile, has problems in the PC realm. Rather than try to become a low-cost leader, the company instead tried to bulk up by buying Compaq Computer. History in the computer market, though, shows that "the key is not scale, the key is low cost," he said in an interview.

    And then later in the article they talk about his positive track record, including his recommendation for HP to buy Compaq:

    While Wall Street analysts have created a cottage industry out of making grandiose (and often ultimately incorrect) predictions and recommendations, Neff can boast of a fairly strong track record of the industry adopting at least some of his ideas. In January 2001, he said that it would behoove HP to purchase Compaq. At the time, most analysts--and even some HP and Compaq execs--warned against buying PC companies, saying it was better to let them fade away.

    So, if he's such a brainiac, why did he think it would be a good idea for HP to buy Compaq, and then call it a blunder after it actually happens.

    It's not a great track record if you recommend something that you end up calling a mistake once it comes true. Bottom line, maybe the world would be a better place if the industry doesn't adopt his ideas.

  • by gmhowell (26755) <gmhowell@gmail.com> on Monday August 05, 2002 @09:12PM (#4015877) Homepage Journal
    Okay, he's a staff writer for 'news.com.com'. What journalistic credit does this guy have? "Hi, I own shares of Dell and Intel. Can I write a 'story' that would pimp their stock prices?" Gimme a break. Perhaps the 50 page report has more info in it, but this is incredibly lame.

    Apple has historically gone to great lengths to be compatible. First they could read PC floppies. Then fat binaries let 68k machines last for a long time after they were no longer sold. There is the compatibility layer in OSX. The idea is simply absurd.

    I know next to nothing about compilers, but doesn't it stand to reason that Apple would have to redevelop most/all of their libraries, to say nothing of the compilers themselves? Particularly if they go off for some 'pseudo-x86' architecture like some are suggesting.

    At that point, what will be the difference between Mac and Windows? Would companies even bother with MacOS ports, or would they just make some bit of middleware, so that the same binary could use the ABI of either system? (I'm talking way beyond my knowledge, so if it sounds like I don't know what I'm talking about, I don't.)

    What would be gained by this? Go from 5% market share to 6%? Not worth the effort. Having access/drivers to PCI/AGP slots, USB, IDE, etc. makes sense. Not for the main architecture.

    Hell, even Transmeta makes more sense than this sort of malarky. Get it to emulate PPC for old apps, ia64 for new stuff, or something like that. But straight Intel hardware? I think not.

    Remember, even though they don't say it, the Mac is the 'computer for the rest of us'. While it's no longer the company line, don't doubt for a minute that Steve likes being a member of the elite. He likes it that cool Hollywood types use iMacs for computer scenes. He likes it that the kids of yuppie hipsters carry iPods.

    Steve is not a commodity guy. Ask the owners of StarMax machines.

    This article (and the one 'proving' the existence of super-duper-top-secret military aircraft) prove that in the eyes of the editors, today was a slow news day. Not slow enough to answer the question "what happens when VA is delisted" but slow, nonetheless.

    • I do not totally disagree with this post but:
      The compiler bit is a little off the only part Apple will have to compile for ia32 (which they do not do already publicly) is the all libraries that go with Mac OS X instead of Darwin, this includes Cocoa, Carbon (since Cocoa is using Carbon for menus and other things) and the window server.

      The kernel is almost compiled fat so is most of the UNIX apps for both ppc and ia32 for Darwin.
      In fact you can compile gcc so it will make fat binaries with one command line.
  • by feldsteins (313201) <scott@scottf e l d s t e i n .net> on Monday August 05, 2002 @09:41PM (#4015995) Homepage
    Apple is quite likely to make a move to another microprocessor. For performance increases, sure. To combat the "megahertz myth," sure. But also so that processors can be had cheaper. For those reasons I think Apple will make the move at some point. The analysts timeframe is probably pretty accurate but I thin it'll be sooner rather than later.

    With that said, there are some concerns. One of them is certainly the size and heat of the current Intel and AMD processors. Fanless iMacs? I don't think so. Could the TiBook have been developed with a P3 or P4? I rather doubt it. That is surely the major drawback that Apple will have to find a way around. They'll either have to engineer their way around it or perhaps Intel et al will start making smaller and cooler products.

    Does anyone back when all PCs used VGA and Apple didn't? Everyone bitched about it but few people actually realized why Apple chose not to use VGA. The reason was that with VGA the computer couldn't auto-sense the display. If it can't auto-sense the display the user has no idea what settings might be valid for it or not, leaving the possibility of choosing invalid ones. (Ever had to answer a trouble call of a user who changed display settings and now the boot sequence ends in a black screen?) That is why Apple didn't use it. One can imagine that they wanted to use it and that they recognized the benefits of using what everyone else used...but that they weren't willing to sacrafice the Apple "ease of use" and "out of box experience" to get it. Interestingly, once comptuers were able to auto-sense the display through VGA connections Apple was right there doing it too.

    Whether one agrees with them on that issue is irrelevant. It's simply important to recognize how Apple thinks about issues like these. Perhaps Apple would really like to use Intel processors but is waiting for resolution to some size and heat issues.

    Just food for thought.

    And incidently, for those who don't know, switching to Intel processors won't mean you can build your own Mac. Forget it. Never happen.
    • ... For performance increases... so that processors can be had cheaper ...

      When Apple picked the PowerPC it was billed as something that would have twice the performance at half the price of the x86. This goal was never realized on the desktop. The fault was not so much a Motorola/IBM failing as it was that no one ever imagined Intel could pull off the absolute miracles necessary to get the x86 to where it is today. The PowerPC is a clean and modern design, easier to work with, etc. but if you can put ten times the effort/money into the x86 then x86 can keep the lead in the desktop arena.
      • Actually they picked ppc because they would own some ip on it and become partners with both MOT and IBM.
        • Actually they picked ppc because they would own some ip on it and become partners with both MOT and IBM

          Let me rephrase: Apple partnered with IBM and Motorola in the PowerPC Consortium because they thought they would eventually have a processor that would offer twice the performance of x86 at half the price. Owning the IP is all about security and investing the resources into PPC was all about forwarding their desktop interests. All this predicated upon the price/performance expectations.
    • Actually, Apple is probably quite unsure of where it's going to go which is why for the first time, they've put their core OS (darwin) out for development on a chip that they don't ship and they're doing it publicly.

      From what I can tell, Motorola's only claim to relevance in AIM is Altivec. If they license it, they can fire a bunch more engineers and chip fabricators to puff up their quarterly results and still collect a check. IBM's got the ability to move things forward and within the next 4 years, I think they'll do it and add altivec to their chip lines. Apple's making it clear that Altivec's not just for the graphics set with their enhancements to BLAST and their commitment to improving GCC support for PPC and Altivec.

      The facts are that for similar chip runs, PPC is cheaper than x86. The chips tend to be simpler and smaller and thus you can fit more on a die with a better error rate.

      The only cost advantage that Intel has is that it is so dominant that it's volume drives the cost down so that it's cheaper. I'm guessing that Apple will stick with PPC or move to Power which might not even require recompiles for most software. If they can innovate themselves into a 8-10% share, the cost difference should go away. At a high enough level, the PPC would be the cheaper choice.
  • Oh, jeez, what a headache if they change.

    Emulation sucks.

    The transition from 68K to PowerPC went better than anyone might have expected, but it was still a headache. As it happened, I was using two Macs at the same time. One was the latest of the 68K generation, the other being the first of the PowerPC generation, and--although it did great on pure-processor benchmarks--the PowerPC was distinctly more sluggish. And crashed more. It really took about two or three years before PowerPC's FELT fast again, and before everyone had native PPC versions of their software.

    I use Virtual PC on my Mac. It works, sort of. For $200-odd it's a great product. It works better than anyone might have imagined, in fact. But it's no substitute for a real PC.

    So what will happen if Apple goes Intel? I assume they'll do their best to provide some kind of PowerPC emulation so that old software will RUN, but I'm sure it will be slow. And buggy.

    And, darn it, old software is IMPORTANT. It's not just a question of the cost of upgrading; I have significant amounts of software that I still use whose companies are either out of business or not upgrading their products.

    And it's always the beloved GAMES that don't run in emulation...

  • 1) Apple needs to sell the total package to make money doing what they do. They do a really good job of it right now IMHO. If I could buy a PC clone and load OS X on it Apple only gets a *very* small portion of that total package. For Apple to make money doing this the cost of OS X would have to be more that even Microsoft Office!

    2) Do you think Microsoft would sell Office for a new "competing" OS? I think they would drop support for OS X in a heartbeat if they did this.

    I don't like to spread rumors of any kind about Apple but I think if they do choose a new CPU it *should* be derived from perhaps the IBM Power series. It has PPC compatibility and is 64bit. Existing software *should* be able to work on it and Apple users would have a lot to look forward to in next generation software.

    The big issue is how much are those mothers gonna cost? In reality I know I don't *need* 64bits to get my work done and that CPU makers are really just cramming it down my throat because they feel the need to sell me something.

    I bought a dual AMD MP 1600 system in the last 6 months and I tend to use my powerbook more often than that [which is only 667 Mhz G4]. I think that says a lot about what I need and what the industry wants me to want.

    I can wait and so can 90% of the public wait for either faster G4's or 64bit Apples.
    • A switch to x86-based hardware does not mean Apple will be killing their hardware business. They do not have to switch to off-the-shelf PC parts. They can continue to use custom and proprietary designs, just substituting an x86 for a PowerPC. They can have the same high standard and reliability. Nothing really changes, x86 Macs are still a different target architecture and MacOS X does not pose any more threat than before to Microsoft.

      One possible exception to the above. Virtual PC's emulation becomes a much more practical option. However Microsoft could buy them out, much simpler than dropping Apple support. Apple support helps with that DOJ monopoly thing.
      • Of course you are right. I think it would be ideal for Apple to do something like use the Crusoe processor core [128 bit VLIW? need to double check] or work on a new unique CPU with AMD or Transmeta or something. Perhaps IBM would want to get in bed too.

        If anything is known about new CPU details Apple is keeping the information in an air-tight container.

  • by jht (5006) on Tuesday August 06, 2002 @07:23AM (#4017237) Homepage Journal
    I don't see this as likely, especially in Neff's timeframe. Here's why. The G4 processor doesn't have the legs that the P4 has right now, but Moto is known to be making at least 1.4 GHz parts right now.

    Apple also has a policy of running duallys at the high end, and given XServe, we know they have a motherboard/chipset in-house that supports reasonably modern features like DDR and ATA-100. And unless all the rumor sites are wrong, there's a new PowerMac due no later than Seybold in about a month - possibly this month.

    So I figure a high-end Mac with dual 1.4 GHz G4 processors, DDR PC2100 RAM, and ATA-100 support is in the cards shortly. That's going to be a reasonably competitive machine for a while, though not quite up to bleeding-edge Wintel specs. There's also likely a little bit more leg in the G4, at least enough to get up around 2 GHz.

    Beyond that, Apple's got some options. They can go to quad processors pretty easily, or by next spring they have a good shot of being up on G5 processors, which are reputedly now in sampling. Should they be making the move to G5, that'll probably carry them another couple of years, so we're talking 2005 at the outside before they have to have the next stop in mind.

    A lot can happen in that time. The likeliest thing is that they jump to a 64-bit contender that emerges by then - possibly AMD but who knows? Migrating to the IBM POWER processors would be another logical move because minimal work would be required and the additional volume would drive IBM's own costs down significantly. Remember, Apple sells more RISC systems in a year than Sun, SGI (though they don't control MIPS anymore), and IBM do combined - yet all those companies see it as worthwhile to continue investing in alternative architectures. If Apple decided to move their volume systems to a slightly scaled-down version of one of these workstation chips it would have a major impact on cost.

    Or Motorola could get serious and start working hand-in-hand with IBM again - IBM's fab capabilities are way beyond Moto's, and IBM could probably build the same G4 as Moto at a higher clock rate with better yields. There is one key reason, though, why Apple doesn't have to worry too much about PowerPC dying - it's huge in the embedded marketplace. Versions of PowerPC are used in all sorts of devices, and I believe it's pretty popular in automotive and networking. That gets your volumes up, too.
  • Speaking of OS ports...

    In the early nineties, one of the knocks on Windows, versus UNIX, was that Windows locked you in to a specific processor architecture.

    When Nt was announced, Microsoft was at great pains to blunt the appeal of UNIX by asserting that NT was highly portable and promising that it would be available on lots and lots of processor architectures.

    I'm not sure I remember all of them, but certainly MIPS, Alpha, and PPC were among them. (Remember the ACE initiative, anyone?)

    All of the versions for non-Intel hardware were late, or had problems, or weren't supported, or never materialized at all. I believe PPC never materialized at all. Alpha never made it past NT 3.5. The promise that NT would be available for multiple processors was pretty much broken in a surprisingly short period of time.

    I keep wondering why this didn't hurt MS in the marketplace. Windows locks you in twice--to Microsoft and to Intel architecture. Admittedly there are viable non-Intel sources for Intel architecture, but still...
    • WinNT 3.1 was initially developed on MIPS and x86. The WinNT 4.0 retail CD sitting on store shelves had x86, MIPS, Alpha, and PowerPC binaries. I recall seeing ads for dual PowerPC 604-120 machines running WinNT 4.0 long before I purchased a PowerMac 8500. Byte magazine reviewed these dual PowerPC machines and showed how they scaled much better than dual Pentium machines.

      Microsoft was entirely successful in delivering cross-platform WinNT up through version 4.0. The problem was that no one purchased the non-x86 machines in signifcant numbers. Nearly everyone preferred low cost and stayed with x86. The few who cared about peformance picked Alpha.

      PowerPC did not have performance and it did not have price. The only thing going for it was the hope of being able to have one machine that could dual boot into WinNT of MacOS. WinNT was basically running on machines built to the PREP spec. A superset of PREP added Apple extensions, this was referred to as CHRP. Solaris, OS/2, and WinNT could run under PREP or CHRP but MacOS required CHRP. Apple kept having delays and MacOS for CHRP missed deadline after deadline. This may have also been about the time Apple began rethinking the Mac clone decision. The reversal on Mac clones may or may not have affected the delivery of MacOS on CRHP. The end result is that without the ability to dual boot to WinNT or MacOS there was little point to WinNT PowerPC. Poor sales led to its demise.
  • by greygent (523713) on Tuesday August 06, 2002 @10:53AM (#4018577) Homepage
    If Apple goes to Intel chips, it doesn't necessarily mean they become PC compatible. There are many other things to an architecture...

    I imagine if they did go to Intel chips they would do something similar to what SGI did with their x86-based machines, and use a custom architecture with a Pentium chip.

    I'm all for Apple going to Intel chips and a custom architecture. I firmly hope that Apple doesn't EVER start making PC compatible machines, and I would wager that if they did, it would lead to their eventual death. I absolutely despise the PC architecture, and aside from OS X, was a major reason for my jump. It's just so... clunky.
  • nVidia (Score:2, Interesting)

    by arloguthrie (318071)
    With all of the rumors going around that Apple may start using nVidia manufactured chipsets, and with nVidia GPUs being as powerful as they are, and with the CEO of nVidia telling WIRED magazine that he wants nVidia to take over CPUs since the bulk of a computer's work for the average user nowadays is rendering the graphics, and with the advent of QuartzExtreme in Jaguar...well, it seems to me that the next manufacturer of CPU's for Apple could very well be nVidia. And then all you gamers could quit whining about Macs. Hell, OS X or OS XI could come with a Cg compiler.

    Hey, it's possible. After all, all we're doing here is throwing around and debating CONJECTURE.
  • I think that most people here are missing the point. I've scanned the discussions, and forgive me if I'm wrong, but everybody is sticking to the argument about speed. This machine is faster, that machine is faster. Apple will do this because of speed, and Apple will do that because of speed. Whoah, whoah! Apple doesn't really care that much about speed!

    Think about it. Sure, they try to ship the newest and the greatest processors when they can, but do you honestly think they'd still be in the AIM partnership if all they cared about was speed? Of course not. The key to understanding Apple is knowing what they value. What do they value? Being the God of their customer's computers.

    Think about it. Apple is constantly building walls between itself and the community. They control all hardware. They are the sole producers of the OS. They approve all drivers. They produce many of the basic Applications one might use (Office Suite, photo program, movie making, burning software, music player, calendar program, scanning software, chatting program, email program). They produce a server that has heavy integration with Macintosh clients. They have a web hosting business that integrates heavily with OS X. The list of internal Apple ties is endless. Sure, you could make the argument that Apple has lots of ties to outside companies and products, but Apple branches out to them (for example, see the digital hub) instead of the companies coming to Apple.

    Apple is building a contained Mac world. They have been forever. Switching to x86 chips would mean losing a lot of control. If they can sacrifice a little bit of speed for a lot of containment, they'll do it in a heartbeat. If you go by Michael Kanellos's stupid argument, Apple will dump their current sound cards and switch to Creative cards within the next couple years as well, because of their better performance. Do you honestly think Apple will want to start relying on another company to produce drivers, tech support, etc.? Apple will produce everything that they can, and when they can't produce it they will invest heavily in a company that can, and set up a strict partnership.

    Earlier I mentioned the AIM partnership. Apple doesn't just buy their chips from the cheapest dealer on the street. They were integrating when that partnership was created, and they continue to integrate today. They won't throw away years and years of work to form a new integration with Intel as a part. It would go completely against Apple's plan.

  • I was reading this http://news.com.com/2100-1001-948493.html?tag=fd_t op article at news.com and it hit me, this could be a great new processor line for Apple. Now befor you mod me as an idiot (this post was made several days after the story aired so I don't think anyone will read this anyway) I don't think that this is apples future chip. I am just pointing out that there are new processors coming out all the time and IBM or Motorola could easily poduce a solution for apple.

To avoid criticism, do nothing, say nothing, be nothing. -- Elbert Hubbard

Working...