Intel shows new chips, outlines platform directions

1356789

Comments

  • Reply 41 of 177
    wmfwmf Posts: 1,164member
    You're forgetting that the inevitable Yonah PowerBook is 32-bit. Face it, OS X Intel is going to start as a 32-bit OS.
  • Reply 42 of 177
    sjksjk Posts: 603member
    Quote:

    Originally posted by Thereubster

    Perhaps what he meant was there is now more logic in Apple producing one, (if they actually can) as there is now a 64bit Intel laptop chip due next year that could be the replacement for a G5.



    Ahh, okay, although for now I bet nearly everyone who see "G5 PowerBook" with insufficient clarification will still think it means the hypothetical PPC-based PB (as I did ).

    Quote:

    I think we will see a G5 Powerbook at Macworld SF in Jan 06.



    A longshot, IMO. However, it seems likely Apple would be designing an Intel-based PB long before the processor transition was announced during the keynote, waiting to finalize as soon as any remaining components (including the processor) were known and available.



    I haven't seen any definitive news about what'll replace Open Firmware. Is the what's-it-called BIOS replacement available now? I thought the specification for it hadn't been finalized yet.
  • Reply 43 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by sjk

    Ahh, okay, although for now I bet nearly everyone who see "G5 PowerBook" with insufficient clarification will still think it means the hypothetical PPC-based PB (as I did ).

    A longshot, IMO. However, it seems likely Apple would be designing an Intel-based PB long before the processor transition was announced during the keynote, waiting to finalize as soon as any remaining components (including the processor) were known and available.



    I haven't seen any definitive news about what'll replace Open Firmware. Is the what's-it-called BIOS replacement available now? I thought the specification for it hadn't been finalized yet.




    Intel calls it EFI. Extensable Firmware Interface. The equivalent to OF which won't work on x86 because several calls aren't supported.



    They haven't had sucess in selling the idea to PC companies, but it's believed that Apple will use it. That might spur PC companies to use it as well. A benefit to Intel of the Mactel relationship.
  • Reply 44 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by wmf

    You're forgetting that the inevitable Yonah PowerBook is 32-bit. Face it, OS X Intel is going to start as a 32-bit OS.



    To whom are you speaking?
  • Reply 45 of 177
    No I really was talking about the low power IBM 970 G5, just to be sure. I know it seems crazy but Apple needs something to hold the Powerbook over until Jan/Feb when Yonah appears. Thats about 6 months away, far too long to let your flagship portable product langish (and no, the G4 7448 will NOT cut it) I'm talking about something aimed at Pro's, not a mainstream product, something that will justify the Powerbooks significant price premium over equivalent PC laptops.
  • Reply 46 of 177
    sjksjk Posts: 603member
    Quote:

    Originally posted by melgross

    Firstly, this won't happen with Leopard. Secondly it's highly unlikely that it would happen with the version after Leopard. Thirdly, Apple won't cut out a vast user base, so until the PPC user base shrinks to a small proportion of the whole, Apple will upgrade the OS to accommodate it. Fourthly, if Apple does use the G5 in a Powerbook (if it's at all possible), that doesn't gaurantee that they will or can do so in an iBook. Fifthly, well there isn't one, it just sounded funky.



    Are we talking about the same thing?



    A OS X xnu kernel compiled exclusively for Intel processors, whether 32 or 64 bit, isn't going to run on any PPC, whether it's 32 or 64 bit. But I'm still trying to fully understand snoopy's last post in the context of this discussion ...
  • Reply 47 of 177
    kickahakickaha Posts: 8,760member
    Quote:

    Originally posted by snoopy

    Well, it would only be a temporary safeguard at most, since most PCs will soon be 64-bit too. It also means that Apple must maintain two separate pre-compiled versions of OS X code, a 64/32-bit for older PPC Macs and a 64-bit only for Intel Macs.



    What, you think they're going to stop shipping PPC upgrades? Feh. Not for three years after the transition, is my guess. Everything will ship as universal binaries for the next few years. Making the Intel side of things 64-bit only is an easy step that doesn't affect the PPC side at all.



    Not saying they'll do it, just that it wouldn't matter one way or the other to owners of PPC Macs in any way, shape, or form if they did.
  • Reply 48 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by sjk

    Are we talking about the same thing?



    A OS X xnu kernel compiled exclusively for Intel processors, whether 32 or 64 bit, isn't going to run on any PPC, whether it's 32 or 64 bit. But I'm still trying to fully understand snoopy's last post in the context of this discussion ...




    Well the trend of the conversation was going in the way of saying that Apple would compile a 64 bit version of the OS that would ONLY work on Intel's 64 bit chips, forcing all of Apple's users to go to those, while rendering all of Apple's current 32 bit machines unusable.



    I'm saying that it won't happen because 32 bit programs won't work on an all 64 bit OS, so Rosetta, which will be in use for years. won't work. Software companies won't re-compile all of their programs to work in an all 64 bit enviornment for years to come either. And Apple will continue to supply a 32/64 bit OS for PPC for years to come as well.
  • Reply 49 of 177
    wmfwmf Posts: 1,164member
    Quote:

    Originally posted by melgross

    To whom are you speaking?



    Sorry. I am criticizing anyone who claims that Apple will/should make OS X Intel 64-bit-only. That doesn't make sense because of Yonah. (BTW, it would be possible for OS X PowerPC to be 32/64-bit while OS X Intel is 64-bit-only, but that still doesn't make sense.)



    BTW, 32-bit programs run great on 64-bit OSes (at least on Windows, Linux, Solaris, AIX, IRIX, etc.). The OS needs to include 32-bit shared libraries, but that's not a problem.
  • Reply 50 of 177
    snoopysnoopy Posts: 1,901member
    Quote:

    Originally posted by Kickaha

    What, you think they're going to stop shipping PPC upgrades? Feh. Not for three years after the transition, is my guess.



    I'm not sure what I said to imply that PPC upgrades will stop. Of course they will continue, and I'm hoping it's for at least 5 years. I'm saying that once Intel PCs are all 64-bit, it will negate the pirating safeguard of a 64-bit-only Mac OS X.



    Quote:

    Everything will ship as universal binaries for the next few years. Making the Intel side of things 64-bit only is an easy step that doesn't affect the PPC side at all. . .



    Okay, it might be easy to do. I didn't know that universal binaries make it unnecessary to maintain separate code for a 64-bit only Mac OS X. However melgross spotted the flaw in this plan. 32-bit PPC code will not run on Rosetta if the OS is 64-bit only.
  • Reply 51 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by wmf

    BTW, 32-bit programs run great on 64-bit OSes (at least on Windows, Linux, Solaris, AIX, IRIX, etc.). The OS needs to include 32-bit shared libraries, but that's not a problem.



    I wish that were so, but it isn't always true. There's the problem also of 64 bit drivers. Right now that's something that Windows 64 is struggling with.
  • Reply 52 of 177
    sjksjk Posts: 603member
    Quote:

    Originally posted by wmf

    BTW, 32-bit programs run great on 64-bit OSes (at least on Windows, Linux, Solaris, AIX, IRIX, etc.). The OS needs to include 32-bit shared libraries, but that's not a problem.



    That's what I thought (with some exceptions, [edit: e.g. drivers]) but according to melgross that won't work with Rosetta. And my original comments were summarized by what Kickaha wrote before that about a 64-bit-only Intel environment not mattering for PPC systems.



    Everything's clearer now, including opinions.
  • Reply 53 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by sjk

    That's what I thought (with some exceptions, [edit: e.g. drivers]) but according to melgross that won't work with Rosetta. And my original comments were summarized by what Kickaha wrote before that about a 64-bit-only Intel environment not mattering for PPC systems.



    Everything's clearer now, including opinions.




    I wish things were clear, but they're really not.



    We don't know how the OS will handle 32 bit 64 bit switching, for example.



    Running 64 bit on the x86 forces the chip to run in 64 bit long mode. Running 32 bit turns 32 bit emulation on. What happens if you're running 32 bit and 64 bit simultaneously? It's not just a matter of librarys you see. The memory model also calls for twice as much memory to be used when running 32 bit apps in a 64 bit enviornment.



    And the driver problem looms large.



    I'd like to know how Apple is going to do this. Will their system work the way it does now 32/64 bit, or will it go just 64 bit? The 32/64 route has advantages for the sort term. No driver problems for example.
  • Reply 54 of 177
    wmfwmf Posts: 1,164member
    Quote:

    Originally posted by melgross

    What happens if you're running 32 bit and 64 bit simultaneously?



    It seems to work fine in Linux, Solaris, etc. (Unless I don't understand what you mean by "simultaneously".)



    Quote:

    The memory model also calls for twice as much memory to be used when running 32 bit apps in a 64 bit enviornment.



    No, it doesn't. Native 64-bit apps suffer pointer bloat, but 32-bit apps use the same amount of memory whether they are running on a 32-bit processor/kernel or a 64-bit processor/kernel.



    In some cases 32-bit apps running on a 64-bit processor/kernel have twice as much address space, but it's hard to imagine why this would cause problems.



    Quote:

    And the driver problem looms large.



    This is indeed a real problem. I wonder if kexts can be fat. If you're building a PPC/x86 kext, how much extra work is it to build PPC/x86/x86-64?
  • Reply 55 of 177
    melgrossmelgross Posts: 33,510member
    Quote:

    Originally posted by wmf

    It seems to work fine in Linux, Solaris, etc. (Unless I don't understand what you mean by "simultaneously".)







    No, it doesn't. Native 64-bit apps suffer pointer bloat, but 32-bit apps use the same amount of memory whether they are running on a 32-bit processor/kernel or a 64-bit processor/kernel.



    In some cases 32-bit apps running on a 64-bit processor/kernel have twice as much address space, but it's hard to imagine why this would cause problems.







    This is indeed a real problem. I wonder if kexts can be fat. If you're building a PPC/x86 kext, how much extra work is it to build PPC/x86/x86-64?




    It depends on how the OS handles it. Some programs cause problems.



    I'll quote from an article in Tom's Hardware. This is the consensis on this.





    "However, the memory advantage can turn into a disadvantage if you don't have enough of it. As each data chunk is 64 bits long, 32 bit chunks of a 32 bit legacy application can consume double the memory compared to running under a 32 bit OS. From this point of view, it does not make much sense to run Windows XP x64 with only a small amount of memory. If you go for this latest version, we recommend installing at least a gigabyte of RAM."
  • Reply 56 of 177
    Quote:

    Factor of 10 is marketing hyperbola. Performance increases by a factor of 10 are "subjective" at best. This plays into their marketing FUD of Performance/Watt FUD.



    Over the next ten years, I think Intel will do a far better job of achieving those goals than PPC.



    Intel/Sony have a Pentium M in a Vaio. Now. And .65 is nearly upon us.



    Apple/IBM don't have a G5 in a laptop now. And do we know when IBM is going .65?



    I don't think 'Performance per watt' is marketing 'FUD' at all. Very few things match the promise when it comes to 'actual' performance (re: broadband throughput scam...) but with 10 of their cpu projects looking towards quad-core while keeping power requirements low...we have a good change of nearing their stated goals.



    I welcome that finally, Intel have got us all out of the 'meghz' race. It means, that, at last, we can now focus on performance, instructions per clock, performance for given watt of power, the architecture of the chip, multitasking, emphasis on software coders becoming more clever and efficient with their program designs to take advantage of multicore cpus.



    It seems coders haven't caught up with the idea of GPU as co-processor and optimising for dual-core/processor systems. SLI isn't optimised yet. PCI-Express doesn't seem any faster.



    It's early days. Chicken and Egg. But as things are not merely getting 'faster' then a new paradigm will emerge.



    I think it may make software more efficient and push the hardware more. As we can't really on Intel 'just making things faster' through pure mhz.



    So, maybe apps like Photoshop are going to have to change to take advantage of dual core and apis like Core Image. Software developers, will have a big part to play in terms of bringing performance advantages. So, Cell maybe hard to program for? But it's a new way of doing things. And learning something new takes time. But judging from what I've seen of the PS3 games...a multi-cpu future coupled with a gpu as co-processor is going to be very exciting when it takes off.



    Lemon Bon Bon
  • Reply 57 of 177
    mdriftmeyermdriftmeyer Posts: 7,503member
    The major change will be the filesystem.



    Back in early 1998 while testing Rhapsody the core engineers worked out the necessities to go from a 32 bit filesystem to 64 bit switch on the fly. The key is once that is done there aint no going back unless you reformat the drive.



    They have been working on moving OS X to 64 bit since late 1997/early 1998.



    If it wasn't for the damn Carbon necessity and legacy crap that came with us ex-NeXT folks merging with Apple we could have seen this transition much sooner.



    On the positive side, a lot of other interesting markets have merged which also refocused Apple's attention to being less interested in scaleable future servers and more around the digital home.



    Seven years and look where we are at. It'll be another 3 years before we really see the transition to 64 bit from the kernel->filesystem->user space /apps is completely accomplished.
  • Reply 58 of 177
    Now that Intel has been forced to change by the laws of physics...we'll start to see the improvements we should have seen forthcoming from the 'promise' of PPC. The market leader who 'owns' the market says things are changing from mhz and...it will. It's an astonishing climbdown. But Apple have just got into bed with the ultimate PPC partner. Cool, slow running chips! Scalable architecture. But, bonus. We get roadmaps. Know what's coming. Plenty of information(!) ((Yeah, 'information' about future product!)) about those plans on wintel web sites like AnandTech etc.



    And, we'll get them much sooner (if the 'lag' of a PPC dual-core/low power G5) is anything to go by. The promise of PPC has been filled with ten years of 'FUD' for the most part.



    604 and G3 being occasional highlights. Some may count the G5 in that.



    Aiming for a 1 watt cpu sounds incredible but...at least it's being aimed for.



    I'd happily take a dual core cpu that can fit in a laptop. It's incredible that it might occur next year?!



    The new cpus from Intel sound great. And we're probably going to get them in Apple kit. You more or less KNOW it because we've seen Intel provide a demonstration of them!



    None of the cloak and dagger stuff of the PPC alliance. Waiting for the G5 (3 gig?) and G6 and Gwhatever have aged me prematurely...



    You'll now have the assurance of knowing what's coming and can plan to buy accordingly.



    Lemon Bon Bon
  • Reply 59 of 177
    Quote:

    Seven years and look where we are at. It'll be another 3 years before we really see the transition to 64 bit from the kernel->filesystem->user space /apps is completely accomplished.



    ...and that fits in nicely with the Intel transition. In three years time we'll wonder what all the fuss was about.



    Apple releases a 'two button' mouse? Yeah, yeah...history man...



    Lemon Bon Bon
  • Reply 60 of 177
    kim kap solkim kap sol Posts: 2,987member
    I hope Steve-o doesn't spoil the Apple/Intel relationship though...



    "By next year, we'll be at 12-core!"



    A year later:



    "I know I was up on stage this time last year promising 12-core Macs but we're not there yet. Introducing the new 8-core PowerBook."



    *Jobs throws the 8-core PowerBook at Otellini and walks off the stage in disgust*



    PS...scramble Otellini's name and what do you get? Intel-lio! Ain't that a little freaky?
Sign In or Register to comment.