2GHz --> 3GHz

1356

Comments

  • Reply 41 of 114
    macjedaimacjedai Posts: 263member
    Quote:

    Originally posted by RBR

    The 980 is the "real deal" even though the 970 is not too shabby, although it seems likely that Apple & IBM will have one 970 speed bump and then the 970+ before the 980 is debuted in a PowerMac. We shall see in time.



    Yup ... Yup ... and Yup ... not sure which three though
  • Reply 42 of 114
    ipeonipeon Posts: 1,122member
    Quote:

    Originally posted by opus

    The new pricing might become necessary once IBM introduces their own Linux-based 970 systems.



    Would MacOS X be able to run on those systems?
  • Reply 43 of 114
    junkyard dawgjunkyard dawg Posts: 2,801member
    I think either,



    Feb 04:

    2.0, 2.2 dual, 2.4 dual.



    Aug 04:

    2.4, 2.6 dual, 3.0 dual.



    Although Apple seems to like doing long product cycles, even when faster CPU are available (although with Moto this is speculation). If they stick to long product cycles, then:



    Aug 04:

    2.0, 2.6 dual, 3.0 dual.



    This is pure speculation, but if IBM is going to "ramp up" 970 speeds on the 130 nm process before migrating to 90 nm, then I'd expect the first scenario with two Powermac speed bumps over the next year. But if IBM fabs 970s at the current speeds and then migrates over to 90 nm for the next speed bump, then it seems likely that well see the second scenario with only a single speed bump. I have no idea how fast the 90 nm transition will happen, but I suspect it will be sometime in spring of next year.



    Still, I get the feeling that I'm thinking about this like it's Motorola, even though IBM blew us away by delivering ahead of schedule and above predicted speeds.
  • Reply 44 of 114
    opusopus Posts: 15member
    "Would MacOS X be able to run on those systems?"



    Rumors of the negotiations with IBM to convince them to commit to the 970 chips, pointed to yes. However, the IBM systems would mainly focused on corporate systems. I beleive the phrasing was something like not systems that directly compete with Apple's current market.



    Frankly, I am not sure. But, growing the MacOS X user base with the help of IBM it not necessarily a bad thing. Microsoft did OK selling mainly OSes. It will recoup a lot of the development cost (~$129/copy). However, Apple has had it very easy with their own hardware and software, they need to be careful, they don't want to spend all their time writing hardware drivers for IBM's 970-based machines.



    I think that IBM and Apple will ensure that they don't step on each others toes here too much. Also, Apple's main competition (and source of new customers) is not going to be MS, but Linux (in its various flavors). MS is scared of Linux and rightfully so. The Linux community is much more openminded about OSes and systems. They are often MS-bashers (which is not a bad thing). Convincing Linux users to switch is liekly an easier option, where Apple can likely pick up a lot of users.



    My question is really how long will MS actively "support" MacOS X. MS office is essential until a MacOS X native MS office compatable option is introduced. You will not get the corporate/business market without MS office, and you may never get those market anyways.



    Opus...
  • Reply 45 of 114
    nebcon65nebcon65 Posts: 47member
    A few points



    First, remember the German IBM site that briefly published specs on an IBM blade using a 2.5Ghz 970? It caused quite a stir on AI for a while.



    Second I don't think that it is necessary for the 970 to go to 90nm in order to put it in a Power Book, I have heard many people that should know that the power draw for a 1.2Ghz 970 would be well within the requirements. i don't think that that is the point however. I think that due to the fact that this is supposed to be the "year of the laptop", that Steve wants to leave 1Ghz far behind on the PB and really knock peoples socks off again with a 1.8 or maybe even a 2Ghz 970 at 90nm. I'm not sure that Apple will ever use the 1.2Ghz part.



    Finally, I wonder with Steve's emphasis on saying the G5 is made for dual processor systems, that they aren't perhaps thinking of stopping at duals and have decided to encourage users to use clustering if they want more. I could be way off here but it's a thought.





    BTW: you may have heard that they caught Baghdad Bob. Perhaps Intel should push for a pardon and hire him for their PR machine.
  • Reply 46 of 114
    banchobancho Posts: 1,517member
    While I do expect IBM to release their own 970 based workstation I would not count on it being a cheap way to get an OS X based machine. Look at IBM's laptops. They are sturdy and reliable (as PC laptops go, but they are also expensive). I'd expect the IBM workstations to go against Sun's workstation lineup.



    I'd love to walk into work one day and see an IBM 970 based workstation in place of my Sparc Ultra 10 though...
  • Reply 47 of 114
    jaredjared Posts: 639member
    I see the next update being,



    2.2



    2.4



    2.6 DP



    then



    2.6



    2.8



    3.0 DP
  • Reply 48 of 114
    boy_analogboy_analog Posts: 315member
    Quote:

    Originally posted by Amorph

    Just wait'll each core is threaded! Then you'll have two processors with two cores each, with each core pretending to be two CPUs.





    Given that hyperthreading sometimes reduces performance (in Intel's implementation, at any rate), I've been wondering about this technology. Would anyone care to guess if threading in 980s will be able to be turned on or off on the fly? Perhaps 980-aware applications will include a preference option to toggle processor threading....
  • Reply 49 of 114
    hmurchisonhmurchison Posts: 12,419member
    Quote:

    Originally posted by boy_analog

    Given that hyperthreading sometimes reduces performance (in Intel's implementation, at any rate), I've been wondering about this technology. Would anyone care to guess if threading in 980s will be able to be turned on or off on the fly? Perhaps 980-aware applications will include a preference option to toggle processor threading....



    IBM believes their version of SMT will be superior to Intels Hyperthreading. %80 efficiency mentioned in a preview article about the POWER5. Hyperthreading surely isn't performing that kind of boost.
  • Reply 50 of 114
    programmerprogrammer Posts: 3,457member
    Did I miss something and IBM has even mentioned the number "980"? Last I heard it was entirely rumour and speculation, but people seem to be treating it an awful lot like fact. The POWER5 is fact, but that is entirely a different animal and it isn't even out yet -- and won't be until next year so assuming that the 980 will arrive shortly thereafter seems aggressive.



    Or did I miss an official roadmap from IBM?
  • Reply 51 of 114
    macjedaimacjedai Posts: 263member
    Quote:

    Originally posted by Programmer

    Did I miss something and IBM has even mentioned the number "980"? Last I heard it was entirely rumour and speculation, but people seem to be treating it an awful lot like fact. The POWER5 is fact, but that is entirely a different animal and it isn't even out yet -- and won't be until next year so assuming that the 980 will arrive shortly thereafter seems aggressive.



    Or did I miss an official roadmap from IBM?




    Nope, ya didn't miss much ... nothing official ... all rumor.



    Did you make it to WWDC?
  • Reply 52 of 114
    Quote:

    Originally posted by Junkyard Dawg



    This is pure speculation, but if IBM is going to "ramp up" 970 speeds on the 130 nm process before migrating to 90 nm, then I'd expect the first scenario with two Powermac speed bumps over the next year. But if IBM fabs 970s at the current speeds and then migrates over to 90 nm for the next speed bump, then it seems likely that well see the second scenario with only a single speed bump. I have no idea how fast the 90 nm transition will happen, but I suspect it will be sometime in spring of next year.



    Still, I get the feeling that I'm thinking about this like it's Motorola, even though IBM blew us away by delivering ahead of schedule and above predicted speeds.




    I get the feeling that the 970 130 nm will not be with us very long. I think IBM will gear up to the 90 nm as soon as they can. I think Apple future hardware depends on the 90 nm 970. PB will not go to the G5 until this happens and Apple servers may not be updated until then. The cooling systems for these products can not compare to the new PowerMacs. We will not see G5 iMac or eMac until the other lines have moved to the G5.
  • Reply 53 of 114
    programmerprogrammer Posts: 3,457member
    Quote:

    Originally posted by MacJedai

    Nope, ya didn't miss much ... nothing official ... all rumor.



    Did you make it to WWDC?




    Nope, they haven't sent me in a very long time.



    There seems to be an awful lot of people assuming an awful lot about the 980, its existance, and the timeframe of its supposed introduction. I suggest we bask in the glow of the 970 for a while and wait for this year's Microprocessor conference in October where we might hear a little more from IBM about their 9xx plans.
  • Reply 54 of 114
    nebcon65nebcon65 Posts: 47member
    Actually Bancho the point I was trying to make was that if IBM had promotional material and specs ready for 2.5Ghz 970 blade 2 months ago it seems that a 2.5 Ghz 970 shouldn't be too far off, and would certainly be ready in 6 months when speed bumped PM's would be expected.
  • Reply 55 of 114
    drboardrboar Posts: 477member
    The 970 needs to get into the iMac soon and then the natrual difference between the iMac and tower would be that all towers would be dual.



    Why iMac need to be 970? The games of course. UT2003 is alredy severly CPU limited on G4s. With games based on the UT2003 Unreal 2 and perhaps Doom III not that far of the G4 has to go.



    At quite anumber of occations the high end iMac has been more or less on CPU parity with the low end tower and for some few hundred more offered both a more advanced optical drive as well as a TFT screen



    Regarding the 2.5 GHz 970. It might be that IBM will skim the cream of the crop for their own blade servers and that is OK by me as the very top CPUs are in small numbers and thus fit a low number product
  • Reply 56 of 114
    mokimoki Posts: 551member
    Quote:

    Originally posted by hmurchison

    Not really. The 970 is basically the interim chip to start the ball rolling. IBM will want to quickly transition to the 980 bing based off the POWER5 Core..it should have internal design tweaks that give it a noticable edge over the 970. The 980 will be 90nm only. This means IBM gets their EF Fab "broken in" with the 970 before moving to the 980. I'm looking forwared to SMT in the 980 giving us the power of Quad processing in a Dual CPU config.



    I disagree. The 970 isn't interim anything, and it should scale quite well, especially with the move to 90nm. I don't expect to see the 980 in anything (ie, a shipping Apple box) until at least 2005, likely 2006.
  • Reply 57 of 114
    Quote:

    Originally posted by moki

    I disagree. The 970 isn't interim anything, and it should scale quite well, especially with the move to 90nm. I don't expect to see the 980 in anything (ie, a shipping Apple box) until at least 2005, likely 2006.



    Sorry for the OT comment: Moki, when is iseek coming out? Sounds pretty cool...



    As for the thread...it just cracks me up that the 970 is just announced and not even shipping, but the crowd is already clamoring for more! I really pity Apple sometimes
  • Reply 58 of 114
    mokimoki Posts: 551member
    Quote:

    Originally posted by Thai Moof

    Sorry for the OT comment: Moki, when is iseek coming out? Sounds pretty cool...



    It should be out in a month or so. Interested in doing the Thai localization?



    I spent some time in Thailand last year actually... beautiful country.
  • Reply 59 of 114
    smirclesmircle Posts: 1,035member
    Going from 2 to 3 Ghz in 12 month sounds impressive - unless you consider that Moores law predicts a doubling of transistor density (translating in a doubling of chip speed) every 18 month.



    IBM does not seem to be able to achieve this. With their predicted rate, they only gain 75% over the next 18 month. As far as I know, intel cpus had seen an even faster gain in clockspeed over the last years.
  • Reply 60 of 114
    telomartelomar Posts: 1,804member
    Quote:

    Originally posted by Smircle

    Going from 2 to 3 Ghz in 12 month sounds impressive - unless you consider that Moores law predicts a doubling of transistor density (translating in a doubling of chip speed) every 18 month.



    IBM does not seem to be able to achieve this. With their predicted rate, they only gain 75% over the next 18 month. As far as I know, intel cpus had seen an even faster gain in clockspeed over the last years.




    Urge to break chairs over people's heads rising...



    How many times does this need to be said:



    Moore's law is is a statement about the economics of integrated circuit manufacture and states absolutely nothing about performance. Transistors do not necessarily directly relate to performance or clock speed.
Sign In or Register to comment.