Faster G4 - MOTO 7470

Posted:
in Future Apple Hardware edited January 2014
«1345678

Comments

  • Reply 1 of 147
    jerombajeromba Posts: 357member
    i want to believe
  • Reply 2 of 147
    thereubsterthereubster Posts: 402member
    .....but deep down you dont..... the register seems to have very high powered RDF generators.......
  • Reply 3 of 147
    addisonaddison Posts: 1,185member
    I believe it. Given the specs of the xServer they could have made an educated guess. In any event I think what they have said is entrely logical and credible.



    There is little point in bringing this increease in speed unless it is coulpled with at least ATA100 support, so I expect that too.
  • Reply 4 of 147
    [quote] The 7470 supports PC2100 DDR-SDRAM memory at 266MHz, as we told you back in February, (see New G4 roadmaps promise Apple harvest

    ), and PC2100 made its debut in Apple's product line with the launch of the Xserve rack last week. More importantly the new G4 supports front side bus speeds of 133MHz and 166MHz, which double to 266MHz and 333MHz.



    The most dramatic bandwidth improvement would be a jump to 266Mhz, and although we don't know if that choice has been made, but a more modest increase to 166Mhz (2x) would be more in keeping with recent announcements. <hr></blockquote>



    THe second paragraph is a bit of a non sequitur since 266Mhz is just a doubled 133 Mhz FSB. A 166 Mhz FSB (which of course would equal DDR333 is actually the more optomistic outcome.



    Still, I remember programmer made reference to something he had seen indicating the G4 would get a 166 bus. Do both have a common source or do they confirm each other? <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />
  • Reply 5 of 147
    Much has been said of the current Mobo failing to feed the current 1 GHz G4 data fast enough to keep it constantly chewing, giving us a hurry up and wait scenario.



    So, my question is, does the speculated new Mobo feed data to the CPU(s) fast enough to keep the CPU(s) constantly fed? Is this another hurry up and wait system?



    Eirik
  • Reply 6 of 147
    xypexype Posts: 672member
    [quote]Originally posted by Eirik Iverson:

    <strong>Is this another hurry up and wait system?</strong><hr></blockquote>



    It's always a hurry up and wait system, but with a faster bus you get less wait. I think the G4 woule be "fed enough" with a 266 FSB - most of the time people wont notice much a difference but for 3D games and "real" work. As programmer pointed out a few times already it will also help the "responsivness" of the system and make the impression that OSX is a lot faster overall.
  • Reply 7 of 147
    stoostoo Posts: 1,490member
    What's so reality distorting about Apple finally moving away from single data rate RAM/FSB? It has to happen sooner or later (hopefully not later <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" /> )



    For those of you wondering, El Reg claims that this FSB is:

    [quote]

    The 7470 supports a modified bus protocol, MPX+, which supports double data transfer and which should effectively run at 266Mhz according to sources.

    <hr></blockquote>



    <a href="http://www.theregister.co.uk/content/3/24018.htmlp"; target="_blank">http://www.theregister.co.uk/content/3/24018.htmlp</A>;



    [ 05-24-2002: Message edited by: Stoo ]</p>
  • Reply 8 of 147
    amorphamorph Posts: 7,112member
    If Apple's MPX+ implementation is as good as their MPX implementation, we should see some serious RAM to CPU bandwidth, which will really make an impact in situations where dual processors and AltiVec kick in.



    Yum. I hope it's true.



    [ 05-24-2002: Message edited by: Amorph ]</p>
  • Reply 9 of 147
    jerombajeromba Posts: 357member
    ok rdf to the max !

    i'm beginning to believe that we will see Dual G4 up to 1.5/1.4 with FSB @ 266 Mhz BUT i'm not hoping for a new case or FireWire 2. I think that all the new 'bells & whistles' will be for the 7500/G5.

    But i hope they can give us a quieter Dual cuz all the QuickSilver are pretty loud.

    40%/50% in speed in one upgrade & MPX+ ...too nice... where is my credit card ???

    54 days and counting
  • Reply 10 of 147
    bigcbigc Posts: 1,224member
    Maybe the 100% speed improvement rumour at xlr8 was correct. July will tell (well probably September by the time they ship, which means Jagwire too).



    Sounds too good to be true, but I'll trade up my dual gigger (but not to apple for $300 though)
  • Reply 11 of 147
    hobbeshobbes Posts: 1,252member
    The Register is a weird mix of ubersceptical and uberoptimistic.
  • Reply 12 of 147
    zazzaz Posts: 177member
    [quote]From the Article:



    The most dramatic bandwidth improvement would be a jump to 266Mhz, and although we don't know if that choice has been made, but a more modest increase to 166Mhz (2x) would be more in keeping with recent announcements.

    <hr></blockquote>



    I think what it means is a top bus speed jump to 266/533mzh is the most drastic jump plausible. The states that a speed of 166/333mzh is more reasonable.



    Regardless, boy, it would be nice. But at this point it is like the picture of a BigMac. Sure, it looked like the greatest burger in the world... so what the hell did i just unwrap?





    The article is really no more than a reasonable speculation based on current PC technology, like many of the posts here.



    Can anyone show me a Register article that was right?
  • Reply 13 of 147
    lemon bon bonlemon bon bon Posts: 2,383member
    <img src="graemlins/hmmm.gif" border="0" alt="[Hmmm]" />



    Lemon Bon Bon
  • Reply 14 of 147
    gamblorgamblor Posts: 446member
    Sorry, but I ain't buyin' it. Why would Apple bother with the new DDR/old MPX motherboard for Xserve if, just two months later, they were going to introduce a new proc/motherboard with DDR & MPX+? Given how much Apple loves to have a single motherboard chipset for everything, I don't really see that happening.



    I think it's much more likely that we'll see PowerMacs with a motherboard similar to the Xserve's, but with more PCI slots, and only two IDE channels. If we ever see a motherboard based on MPX+, it's not going to show up until MWSF at the earliest, and probably MWNY '03.
  • Reply 15 of 147
    programmerprogrammer Posts: 3,409member
    [quote]Originally posted by Gamblor:

    <strong>Sorry, but I ain't buyin' it. Why would Apple bother with the new DDR/old MPX motherboard for Xserve if, just two months later, they were going to introduce a new proc/motherboard with DDR & MPX+? Given how much Apple loves to have a single motherboard chipset for everything, I don't really see that happening.



    I think it's much more likely that we'll see PowerMacs with a motherboard similar to the Xserve's, but with more PCI slots, and only two IDE channels. If we ever see a motherboard based on MPX+, it's not going to show up until MWSF at the earliest, and probably MWNY '03.</strong><hr></blockquote>



    Actually, if done right then the Xserve chipset already supports DDR and it was just waiting on the 7470. Since the 7470 is a new chip and likely has a constrained supply, Apple may want to focus on making the PowerMacs competitive rather than make the Xserve faster when it doesn't really need to be. They may also apply this going forward -- give the iMac the same chipset, but the 7455 instead of the 7470 which would give a clear performance difference between the two lines of machines.
  • Reply 16 of 147
    stoostoo Posts: 1,490member
    Doh! Programmer beat me to it
  • Reply 17 of 147
    spookyspooky Posts: 504member
    way beyond . . .
  • Reply 18 of 147
    brussellbrussell Posts: 9,812member
    Oh, jeez, this had to come and ruin my day. The Register has just been so utterly, absolutely, fantabulistically wrong about new Mac processors, that no matter how right this sounds, it can't be true.
  • Reply 19 of 147
    gamblorgamblor Posts: 446member
    [quote] Actually, if done right then the Xserve chipset already supports DDR and it was just waiting on the 7470. Since the 7470 is a new chip and likely has a constrained supply, Apple may want to focus on making the PowerMacs competitive rather than make the Xserve faster when it doesn't really need to be. They may also apply this going forward -- give the iMac the same chipset, but the 7455 instead of the 7470 which would give a clear performance difference between the two lines of machines. <hr></blockquote>



    Unfortunately, that's a pretty damn big "if", and there's no hard and fast evidence to support it. I've seen posts (mainly on Ars) from a few people "in the know" (primarily MrNSX & BadAndy) who've stated that redesigning the MPX bus to support DDR would be a monumental task. I'd certainly like to think it's true, though.



    By the way, what evidence have we got that the 7470 is a real chip, other than a few articles at the Register? Remember, these were the same idiots who were convinced we'd see the G5 (8500) at this past MWSF. What did we get? The 7455. Not to say the 7455 isn't a great chip (in fact, Moto's progress during 2001 was pretty damn good), but it ain't no G5.
  • Reply 20 of 147
    programmerprogrammer Posts: 3,409member
    [quote]Originally posted by Gamblor:

    <strong>



    Unfortunately, that's a pretty damn big "if", and there's no hard and fast evidence to support it. I've seen posts (mainly on Ars) from a few people "in the know" (primarily MrNSX & BadAndy) who've stated that redesigning the MPX bus to support DDR would be a monumental task. I'd certainly like to think it's true, though.



    By the way, what evidence have we got that the 7470 is a real chip, other than a few articles at the Register? Remember, these were the same idiots who were convinced we'd see the G5 (8500) at this past MWSF. What did we get? The 7455. Not to say the 7455 isn't a great chip (in fact, Moto's progress during 2001 was pretty damn good), but it ain't no G5.</strong><hr></blockquote>



    The "if" I postulated is perfectly likely -- that Apple built DDR into their chipset. The much less likely "if" is whether the 7470 exists and supports a DDR MPX FSB. I agree that that is quite a big "if". The minimum possibility is an Xserve-in-PowerMac-clothing because we know that they already have that technology. The wild optimists are hoping for a RapidIO-based 7500. The slightly more realistic types are hoping for the 7470. Won't know what we'll really get until July 16.
Sign In or Register to comment.