AltiVec implementation in the PPC G5

13»

Comments

  • Reply 41 of 51
    [quote]Originally posted by Rmh1572:

    <strong>Hey ada I like your short crisp answer but can you elaborate on how you know or if it is just a guess. if not well then mystery will prevail</strong><hr></blockquote>



    rmh, i wish i could. but i really can't elaborate too much. the g5 is going into production in enough time for them to be out in july. since i can't really say where i get my info, take it as you will. i'll just let you know that i am excited about it coming out mid year. i don't know the specs of the chip, just the timing of it. sorry.



    [quote]What is this bullshit about "someone else" delivering the next generation Gx chip? LOL! What a load of crap! Quite fankly, Motorola and IBM are about neck -and-neck when it comes to real breakthroughs in processor design. Motorola had a disasterous incident with the cache-coherency problem with the G4, and even more serious fab problems, but this doesn't preclude Motorola from taking the lead in CPU design again.<hr></blockquote>



    i'm not saying that moto CAN'T make it. they had a project that was nixed. i'm saying they're NOT doing it.



    [ 02-27-2002: Message edited by: admactanium ]</p>
  • Reply 42 of 51
    onlookeronlooker Posts: 5,252member
    [quote]Originally posted by AirSluf:

    <strong> <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />



    Man is that a screwed up and completely wrong statement.



    </strong><hr></blockquote>



    Well yes it is when ou take it out of it's context as you just did by only displaying half of the paragraph.



    [quote]Originally posted by Rmh1572:

    <strong>Hey ada I like your short crisp answer but can you elaborate on how you know or if it is just a guess. if not well then mystery will prevail</strong><hr></blockquote>



    There you go, the mystery will prevail. I never said it would, or would not have it. I said as I recalled it was in motorolla's PPC roadmap originaly, and then was removed sometime after IBM had said (something like) they were not including it in their version of the G5 that they had. They didn't think it was all that impressive at the time apparently.



    But before all this started I wanted to know where all these people were getting this G5 Altivec information from. I have yet to see one scrap of spacific Altivec info online anywhere reguarding the G5, amd there are like 10 people here saying "I heard it had this, and well it's supposed to have this, and it will have double the that", which is what I am interested in. I would like to read some of these articles myself. I can't find them.

    I must be stupid. :eek:
  • Reply 43 of 51
    Does anybody seriously doubt that whatever chip Apple calls the G5 won't have altivec? Look at how much effort Apple has put into optimizing their code for it-- iTunes, iDVD, iMovie, iPhoto all make use of altivec. OS X uses altivec extensively for openGL acceleration, and for Quartz. Not to mention Final Cut Pro and DVD Studio Pro. Look at Apple's recent work with optimizing libraries, like the ones used by the Stanford researchers for gene sequencing (that was what it was for, right?)



    Look at all the work Apple has done optimizing for altivec up to this point, and there's one inescapable conclusion: Apple is wed to altivec. Any new processors they consider are going to have to use it.
  • Reply 44 of 51
    msleemslee Posts: 143member
    Gamblor...maybe you should spend more time in the BF...heh



    BLAST is a database search that allows you to do sequence similarity searches for polypeptides and polynucleotides.
  • Reply 45 of 51
    powerdocpowerdoc Posts: 8,123member
    [QUOTE



    Look at all the work Apple has done optimizing for altivec up to this point, and there's one inescapable conclusion: Apple is wed to altivec. Any new processors they consider are going to have to use it.[/QB][/QUOTE]

    Yes if the G5 from IBM is ship without altivec or the like, they won't go in the Mac.
  • Reply 46 of 51
    onlookeronlooker Posts: 5,252member
    [quote]Originally posted by Gamblor:

    <strong>Does anybody seriously doubt that whatever chip Apple calls the G5 won't have altivec? Look at how much effort Apple has put into optimizing their code for it-- iTunes, iDVD, iMovie, iPhoto all make use of altivec. OS X uses altivec extensively for openGL acceleration, and for Quartz. Not to mention Final Cut Pro and DVD Studio Pro. Look at Apple's recent work with optimizing libraries, like the ones used by the Stanford researchers for gene sequencing (that was what it was for, right?)



    Look at all the work Apple has done optimizing for altivec up to this point, and there's one inescapable conclusion: Apple is wed to altivec. Any new processors they consider are going to have to use it.</strong><hr></blockquote>



    Optimisim is a good thing. And I too think that Apple would be in the crapper without Altivec. Now that there is a thread mentioning a 4GHz Pentium (that I have yet to read) I would suspect Apples plans for processors on their next upgrade will be to the tune of 2GHz DP's at the least. I would hope. I stll want a G5 now that I'm pretty sure it will have Altivec.
  • Reply 47 of 51
    airslufairsluf Posts: 1,861member
  • Reply 48 of 51
    onlookeronlooker Posts: 5,252member
    [quote]Originally posted by AirSluf:

    <strong> <img src="graemlins/bugeye.gif" border="0" alt="[Skeptical]" />



    Man is that a screwed up and completely wrong statement.



    IBM has NEVER had any intention of using Altivec in ANY of their processors since Mot pulled licensing games during the G3 era before Mot even delivered a finished Altivec unit. IBM also doesn't put the IBM roadmap on Mot's site, they have enough server space and sense to keep it on their own. It also says they are including a SIMD unit on their next generation desktop chips. Not Altivec, but same principle. Gets them around the licensing issues of Mot's implementation of the jointly owned standard.</strong><hr></blockquote>



    No airhole that's a totaly srewed up and wrong statement.



    My bad was to use the word paragraph instead of post. You took half of the post, and replaced it's context with your misrepresentation of what I said as if your misrepresentation was what I said.



    Your mistake not mine.



    the following was posted by Airsluf

    [quote]IBM has NEVER had any intention of using Altivec in ANY of their processors since Mot pulled licensing games during the G3 era before Mot even delivered a finished Altivec unit.<hr></blockquote>



    I presume that is just a statement of yours.



    It touches on a something I said from a more recent era, but does not seem spacificaly related to my post other than that it touches on what IBM was doing, or intending to do at a time before the period I was refering to.



    the following was posted by airsluf

    [quote]IBM also doesn't put the IBM roadmap on Mot's site, they have enough server space and sense to keep it on their own<hr></blockquote>



    I think this part is funny because I have no idea what this has to do with my post. I never said IBM's roadmap was posted on mot.com nor did I mention IBM's roadmap at all. Actually I had no idea they (IBM) posted their own roadmap for the PPC.



    \\-Although, I am going to go check it out in a minute now.-/



    What I did say was something similar to: IBM stated, or I may have said: IBM said: they were not intending on using Altivec in the version of a G5 that they had planned. Give or take a few words as I recall...



    Wait a second. Let me put the last part up.



    Ok I assume this is refering to IBM's roadmap at their site.



    More from airsluf

    [quote]It also says they are including a SIMD unit on their next generation desktop chips. Not Altivec, but same principle. Gets them around the licensing issues of Mot's implementation of the jointly owned standard.<hr></blockquote>



    General information, but nothing I desputed in my post nor did I refer to it.



    Now that look at it all. your mighty battle rant has very little (nothing spacific) to do with what said in my post



    [Edit drunken rants deleted]



    [ 02-28-2002: Message edited by: onlooker ]</p>
  • Reply 49 of 51
    airslufairsluf Posts: 1,861member
  • Reply 50 of 51
    telomartelomar Posts: 1,804member
    onlooker that petty display was unnecessary. There was no need to stoop as low as you did <img src="graemlins/oyvey.gif" border="0" alt="[No]" />



    Edit: Double word removed



    [ 02-28-2002: Message edited by: Telomar ]</p>
  • Reply 51 of 51
    onlookeronlooker Posts: 5,252member
    [quote]Originally posted by Telomar:

    <strong>onlooker that petty display was unnecessary. There was no need to stoop as low as you did <img src="graemlins/oyvey.gif" border="0" alt="[No]" />



    Edit: Double word removed

    </strong><hr></blockquote>



    Yea, I know.
Sign In or Register to comment.