Macbidouille: 970's on sale at the end of May

1910121415

Comments

  • Reply 221 of 300
    jcgjcg Posts: 777member
    Quote:

    Originally posted by Nevyn

    Depends on what "is" is. Er, wait.





    So iCal just needs "64-bit sane" to go 2x as fast.

    Photoshop will go 2x at "64-bit sane"...

    ...but be able to access more memory/plugings at "64-bit aware"

    Oracle will see a tremendous boost (Greater than 2x) at "aware"



    But for all the nitty-gritty things that might see a _little_ improvement, that's going to wait for Apple to put out the "full 64-bit OS".





    Will iCal really be twice as fast on a 970 with a 64 bit OS as it would on a 32 bit OS on the same hardware? I thought I read that 64 bit might actually have a slight depreciation in speed for some tasks which dont require 64 "bitness", say adding two real numbers such as 1+1 becouse the registers (?) for the 64 bit version are twice as long to desciribe "1". Wasnt this one of the reasons that Apple/IBM went with a PPC which was fully backward compatable with the 32 bit code?
  • Reply 222 of 300
    programmerprogrammer Posts: 3,458member
    Quote:

    Originally posted by Nevyn

    You aren't being obtuse, this is confusing, and I still don't think I explained it well... but I have to run for now.



    Hmmm... how about we call these two kinds of support:



    a) 970-enabled OS

    b) 64-bit OS



    If the 970 hardware arrives before Panther there will be a 970-enabled Jaguar update that is not a 64-bit OS. Panther will be a 64-bit OS and 970-enabled.
  • Reply 223 of 300
    programmerprogrammer Posts: 3,458member
    Quote:

    Originally posted by JCG

    Will iCal really be twice as fast on a 970 with a 64 bit OS as it would on a 32 bit OS on the same hardware? I thought I read that 64 bit might actually have a slight depreciation in speed for some tasks which dont require 64 "bitness", say adding two real numbers such as 1+1 becouse the registers (?) for the 64 bit version are twice as long to desciribe "1". Wasnt this one of the reasons that Apple/IBM went with a PPC which was fully backward compatable with the 32 bit code?



    No, iCal would be slightly slower if it were a 64-bit app (comparing 64-bit app on 970 to a 32-bit app on 970, assuming both are optimized for the 970).



    A 64-bit app will not be faster than a 32-bit app unless it is doing things which are enabled by having a 64-bit machine -- i.e. 64-bit integer arithmetic and 64-bit memory accesses.
  • Reply 224 of 300
    jcgjcg Posts: 777member
    Quote:

    Originally posted by Programmer

    No, iCal would be slightly slower if it were a 64-bit app (comparing 64-bit app on 970 to a 32-bit app on 970, assuming both are optimized for the 970).



    A 64-bit app will not be faster than a 32-bit app unless it is doing things which are enabled by having a 64-bit machine -- i.e. 64-bit integer arithmetic and 64-bit memory accesses.




    Thank you Programmer for the clarification.
  • Reply 225 of 300
    nevynnevyn Posts: 360member
    Quote:

    Originally posted by JCG

    Will iCal really be twice as fast on a 970 with a 64 bit OS as it would on a 32 bit OS on the same hardware?



    Um. No. Sorry. That should be "Twice as fast compared to a G4". I concur with Programmer, if anything iCal will run slower as a 64-bit app on 64-bit hardware compared to as a 32-bit app on the identical hardware.
  • Reply 226 of 300
    shetlineshetline Posts: 4,695member
    Quote:

    Originally posted by Programmer

    If the 970 hardware arrives before Panther there will be a 970-enabled Jaguar update that is not a 64-bit OS.



    I suspect that this pre-Panther OS update would not have to be anything more than a few changes needed to duly recognize the presence of a 970, and then make sure that the 970 was placed into, and stayed firmly locked into, its 32-bit mode?
  • Reply 227 of 300
    programmerprogrammer Posts: 3,458member
    Quote:

    Originally posted by shetline

    I suspect that this pre-Panther OS update would not have to be anything more than a few changes needed to duly recognize the presence of a 970, and then make sure that the 970 was placed into, and stayed firmly locked into, its 32-bit mode?



    The format of the virtual memory page tables is (probably) changed to accomodate 64-bit address spaces. Even if large spaces are not supported the kernel would need to be changed slightly to handle the new format. The supervisor instructions and registers for the 970 are (probably) different than in the Motorola processors as well, requiring changes to other parts of the kernel and possibly some drivers. The motherboard will have a new suite of devices (AGP 8x, new PCI, USB 2, new memory subsystem, possibly VSPs) requiring driver updates. Many of these are similar to other new hardware releases, but the 970 changes will be just a bit more than that. Certainly nothing is stopping Apple from calling it a 10.2.x release, however.



    10.3 will probably be more widely recompiled to be optimized for the 970, however, and will see significantly improved performance on the new hardware in addition to having 64-bit support.
  • Reply 228 of 300
    rhumgodrhumgod Posts: 1,289member
    Quote:

    Originally posted by Nevyn

    Um. No. Sorry. That should be "Twice as fast compared to a G4". I concur with Programmer, if anything iCal will run slower as a 64-bit app on 64-bit hardware compared to as a 32-bit app on the identical hardware.



    Frameworks (APIs) can be 64-bit enabled, which allow programs to run in the 64-bit "framework" of the OS. While the hardware may be 64-bit, remember that the 970 is backward compatible with 32-bit OSs and software. That is how Apple is going to be able to get away with releasing 64-bit hardware running 32-bit OS and apps initially. Think of it this way: You buy an HDTV-capable television set. Granted, the HDTV decoder is not included so you are not watching HDTV broadcasts in the true HDTV format, but are capable of doing so with the proper decoder. Same goes for the 970 Powermac. It is 64-bit hardware, but you will initially be running 32-bit OS and apps.



    While I agree with people who have said a 64-bit app will not be twice as fast as a 32-bit app, the newer hardware will seem like it is twice as fast because of the inherent improvements in moving information around various hardware subsystems (ie; memory, buses, cpu, fpu, etc...).



    Regardless, this thing is going to scream!
  • Reply 229 of 300
    tsukuritetsukurite Posts: 192member
    Quote:

    Originally posted by kraig911

    i'm going to say they are wrong... in which case I hope I am, but everytime I make a prediction it has never become right... so I'm hoping my bad luck will give me good luck so that this whole 970 happens now... so all I have to worry about is the money to get one but with my luck I will finally be correct but in this case I don't want too.



    Wow, after reading this, my head hurts!
  • Reply 230 of 300
    rhumgodrhumgod Posts: 1,289member
    Quote:

    Originally posted by tsukurite

    Wow, after reading this, my head hurts!



    Grammar is becoming a lost art, alas.
  • Reply 231 of 300
    Quote:

    Originally posted by Rhumgod

    Grammar is becoming a lost art, alas.



    [irony alert]

    Unfortunately, that is the state our society is at.

    [/ia]

  • Reply 232 of 300
    programmerprogrammer Posts: 3,458member
    Quote:

    Originally posted by Rhumgod

    While I agree with people who have said a 64-bit app will not be twice as fast as a 32-bit app, the newer hardware will seem like it is twice as fast because of the inherent improvements in moving information around various hardware subsystems (ie; memory, buses, cpu, fpu, etc...).



    32-bit apps running on this hardware will gain these same benefits, however. The 64-bit and 32-bit apps will have no significant performance differences unless they are working on 64-bit integers, or require >32-bit address space. There is no question that the 970 will be faster than a G4, but this has nothing to do with 64-bit vs. 32-bit -- it has to do with a modern very superscalar, heavily pipelined out-of-order-execution core with gobs of bandwidth available.
  • Reply 233 of 300
    bartobarto Posts: 2,246member
    Ok, getting back to what's really important here. How kewl will the next Power Mac look, and what is the slogan going to be? It is AppleInsider after all!



    Illustrator is FUN. I kinda hit a brick wall (in terms of functionality) doing mock-ups in AppleWorks Drawing. Some of you may remember my stackable Mac. What do you guys think is better:



    Power Mac 9700: A New Golden Age







    Power Mac 9700: Back in Black







    Hmmmmm?



    Barto
  • Reply 234 of 300
    rhumgodrhumgod Posts: 1,289member
    I think the system board is going to be longer and narrower as the CeBIT image that "snuck out", and several other recent rumor sites have noted. I am guessing a bit deeper of a system then. I won't even try and work up a mockup as Apple engineers think way ahead of existing or even prototype systems.
  • Reply 235 of 300
    kroehlkroehl Posts: 164member
    Quote:

    Originally posted by MartianMatt

    [irony alert]

    Unfortunately, that is the state our society is at.

    [/ia]





    <nitpick>

    You are IN a state - not at one. So that would be:



    'Unfortunately, that is the state our society is in.'



    </nitpick>







    <edit: forgot all the smilie thingies>



    Kroehl
  • Reply 236 of 300
    macroninmacronin Posts: 1,174member
    Quote:

    Originally posted by Messiah

    Why is there all this excitement about moving to a 64bit OS on a 64bit CPU, when clearly it's not going to run any faster than a 32bit OS on a 64bit CPU?



    It's all about getting our 64bit goodness out the door & onto the desktop/laptop before Wintel...



    ;^p
  • Reply 237 of 300
    Quote:

    Originally posted by kroehl

    <nitpick>

    You are IN a state - not at one. So that would be:



    'Unfortunately, that is the state our society is in.'



    </nitpick>







    <edit: forgot all the smilie thingies>



    Kroehl




    Gotcha!



    Yup, that was the point of the [irony alert]...



  • Reply 238 of 300
    mmicistmmicist Posts: 214member
    Quote:

    Originally posted by kroehl

    <nitpick>



    quote:

    Originally posted by MartianMatt

    [irony alert]

    Unfortunately, that is the state our society is at.

    [/ia]





    <nitpick>

    You are IN a state - not at one. So that would be:



    'Unfortunately, that is the state our society is in.'



    </nitpick>





    Kroehl




    Ending sentences with prepositons, is, of course, grammatically wrong.



    Should be : Unfortunately, that is the state in which our society is.



    Churchill showed just how silly this can get with the marvellous phrase "This is something up with which I shall not put."



  • Reply 239 of 300
    amorphamorph Posts: 7,112member
    This is the state in which our society is at?
  • Reply 240 of 300
    kurtkurt Posts: 225member
    I may have just missed his posts but I haven't seen Moki on these boards for a while. No comments on the MB rumors for dates and benchmarks. You don't suppose he has a test machine and is under a NDA do you?



    It's fun to start rumors.
Sign In or Register to comment.