CONFIRMED: MPC 7457 with up to 1833 Mhz

2456789

Comments

  • Reply 21 of 173
    [quote]Originally posted by Tomb of the Unknown:

    <strong>

    Wrong again. It's estimated result is under 1000, there are no actual benchmark numbers for this chip yet.



    So out of your first three assertions (the facts of which are a matter of public record at this point) you get exactly zero right and we're supposed to believe the rest of your drivel?



    Here's a thought. Next time you have a burning desire to share what I'm sure you regard as burning insights into the semiconductor industry, stop and ask yourself if you really know what you're talking about.



    If the answer is no, then don't post, 'k?</strong><hr></blockquote>

    As much as I'd like to agree with most of what you have said...here are some business practices for you to go and ponder.

    And for u who don't know, I for one am excited about this chip too!



    Yes...We don't know for sure when they are coming out, but believe me, there is no reason to assume it was possibly coming out in January--you would be sure to know that if IBM had any chance of doing that or anywhere near that they would have said something more like, "Somewhere in the first 1/2 of 2003", instead of "End of 2003, subject to change" That subject to change, 10 bucks says, and implies LATER than mid-2003.



    The clock rate they say will go to around 1800, a safe bet that would be the max or mot even reach that high at first (which you noted.)



    And for the SPEC rumors, believe me, and any other business--that they are going to "Assume" as high as they can without seeming like liars and untrusting to the public, the whole reason for the announcement in October is purely PR, so you know they are trying their best to market this chip.



    So out of your points I would also assume you get no points. This is all assumption, although I think I have logic on my side, logic isn't always right. And try to be considerate next time to posters, I'm sure you've been wrong before.
  • Reply 22 of 173
    matsumatsu Posts: 6,558member
    I think it's still possible that we may see next gen chips from both IBM and Moto.



    IBM at least has its cards on the table, and while they have sucked harder than even Mot in the last 3 years (600-800Mhz G3's, come on!) they at least are showing something promising, talking about it publicly, aiming it at the desktop market (rather than embedded) and giving real timelines. Moto, for whatever reasons, is not -- maybe it's politics and maybe they just don't have anything, you can't really say. I'm inclined to think Moto did the work to develop a new chip, but they may not apply whatever designs to an Apple-pleasing chip unless they have a reason to do so. (ie, commitment to use it)



    Otherwise, a .13u G4 with a decent FSB, more L2, more support for L3, and low power consumption, might be aimed primarily at the embedded markets, but it might still be something that's good for notebooks or even the tablets that have most of you all moist in the panties.



    Assume G5 refers to next generation, I think it possible that we may see a G5pro from IBM and a G5light/G4.5 from Moto. The former will be 64bit, big FSB, all the goodies, etc etc.. the latter may be either a 32bit enhanced G4 (die shrink, cache enhancements, FSB improvements) and progress to something 8xxx (BookE ??) based, perhaps not even with much greater performance, but also 64bit in an effort to have maximum buzzword compliance.



    [ 10-22-2002: Message edited by: Matsu ]</p>
  • Reply 23 of 173
    [quote]Originally posted by robster:

    <strong>



    Dude, there is NO desktop G5, never will be!

    Your comparing an existing, announced IBM chip to a non-existent chip from a company who are looking to sell - off there chip division.

    you ought to THINK a bit first.

    Sorry to flame but your so wrong it's just not funny...

    It's a bitch that we're stuck with the damn G4 for ages until but it won't be as long as you think, IBM say the chip is available in Q3 2003 but that's only for normal customers, Apple are a development partner and will have it in quantity end of Q1 2003.



    Whoa end of flame!</strong><hr></blockquote>

    What makes you think Apple will have it Q1 2003?

    All we know of is Q3...?
  • Reply 24 of 173
    "The lines I liked involved things like: 'First of a new chip family', or 'These numbers are conservative', or 'The Fishkill plant will be running full tilt until Summer 2003, roughly when we intend to transfer to 0.90 nm'. "



    Sounds enticing...almost teaseworthy. I'd like to think IBM has an Ace up their sleeve.



    With the 'Vmx' aspect, it seems Apple was at least considered in the development of this chip. Maybe Apple will have 'special' early intro' privelages.



    Optimist in me says Apple go G5 in Easter. Realistic? July intro' with later shipping. The pessimist? Let's not go there.



    We'll have to wait and see.



    Lemon BOn BOn
  • Reply 25 of 173
    outsideroutsider Posts: 6,008member
    True. Apple was so early in getting the G4 7400 they barely had enough to ship at 450MHz...
  • Reply 26 of 173
    sc_marktsc_markt Posts: 1,393member
    [quote]Originally posted by rickag:

    <strong>



    The MPC75XX is just that, a rumor. An old, very very very old rumor.



    </strong><hr></blockquote>



    Tell that to Motorola. Somebody here posted (I don't remember which thread) a link to a Motorola web site and on the site, they had a reference to the MPC75XX.



    If I find the link, I'll post it.
  • Reply 27 of 173
    sc_marktsc_markt Posts: 1,393member
    Rickag.



    Here are the Mototola links that have reference to the 75XX parts:



    <a href="http://e-www.motorola.com/brdata/PDFDB/docs/MPC7450CE.pdf"; target="_blank">http://e-www.motorola.com/brdata/PDFDB/docs/MPC7450CE.pdf</a>;



    <a href="http://e-www.motorola.com/brdata/PDFDB/docs/MPC7455CE.pdf"; target="_blank">http://e-www.motorola.com/brdata/PDFDB/docs/MPC7455CE.pdf</a>;



    Just download the PDF's, open them and do a search for 75xx.
  • Reply 28 of 173
    rickagrickag Posts: 1,626member
    [quote]Originally posted by sc_markt:

    <strong>Rickag.



    Here are the Mototola links that have reference to the 75XX parts:



    <a href="http://e-www.motorola.com/brdata/PDFDB/docs/MPC7450CE.pdf"; target="_blank">http://e-www.motorola.com/brdata/PDFDB/docs/MPC7450CE.pdf</a>;



    <a href="http://e-www.motorola.com/brdata/PDFDB/docs/MPC7455CE.pdf"; target="_blank">http://e-www.motorola.com/brdata/PDFDB/docs/MPC7455CE.pdf</a>;



    Just download the PDF's, open them and do a search for 75xx.</strong><hr></blockquote>



    I stand corrected. Promote the 75XX from rumor to either vaporware or a typo







    Just thought I'd add that I believe Motorola is and has been working on new and better processors. What the numbering scheme is, I don't know or care.



    Will they introduce a processor with Rapid I/O with a similar set of logic units present in the G4(integer, dp floating point, SIMID, etc.)? Certainly they will, other customers want this or will, this is a no brainer.



    The question is when, and will they satisfy Apple at that time? By the time they introduce a processor like this, it may only be suitable for the embedded market, who knows. They needed to introduce faster G4's with higher bandwith last January, not January 2004.



    Apple has been able to ride the Motorola embedded wave for a few years now, but it may be over. A MPC7500 in a PowerMac would be great, six months ago. In 2004, too little too late for the Powermac's. Maybe laptops, will see.



    Maybe I'll be floored early in 2003 and the mythical 75XX's will appear, but I really doubt it. I fully expect Motorola's original Apollo, the one with HiP7, to appear in early 2003, on the same MPX bus @ 167MHz., probably duals. But no Rapid I/O, no frontside bus capable of DDR.



    Motorola didn't have the resources or the kahones to invest in a cpu designed specifically for desktops. Fool's bet, maybe, but look @ what Apple has, OS X on UNIX(unreal when you think about it really) digital hub strategy(brilliant), LCD monitors pretty much across the line, built in firewire, fast ethernet built in, airport, etc.



    I think Motorola lost the bet against Apple and in the long run may regret it. Aren't they considering selling off the semiconductor part of their business.



    IBM now is very much interested in the desktoop market or why the 970. Is IBM betting on Apple and putting their money where it counts. We'll see.



    [ 10-22-2002: Message edited by: rickag ]



    [ 10-22-2002: Message edited by: rickag ]</p>
  • Reply 29 of 173
    henriokhenriok Posts: 537member
    [quote]Originally posted by Programmer:

    <strong>I see a lot of comments that people are convinced that the current G4 7455s are the Apollo. Is there evidence to that or is just an assumption?</strong><hr></blockquote>



    When doing a search for "apollo" on Motorolas site you'll get a lot of links to documentation for 7455, like this one: <a href="http://e-www.motorola.com/webapp/sps/site/prod_summary.jsp?code=MPC7455&nodeId=01M98653"; target="_blank">MPC7455 Product Summary Page

    </a>.

    On that page there is reference to "MPC7455 Rev2.0 (Apollo), 483 CBGA, 2.5V Bus, 2.5V L3".



    Among the menues in the

    <a href="http://e-www.motorola.com/cgi-bin/faq.cgi?kbase=powerpc"; target="_blank">FAQ</a> the 7445/7455 is called Apollo, and we see the codenames of other processors too.
  • Reply 30 of 173
    rickagrickag Posts: 1,626member
    [quote]Originally posted by Henriok:

    <strong>



    When doing a search for "apollo" on Motorolas site you'll get a lot of links to documentation for 7455, like this one: <a href="http://e-www.motorola.com/webapp/sps/site/prod_summary.jsp?code=MPC7455&nodeId=01M98653"; target="_blank">MPC7455 Product Summary Page

    </a>.

    On that page there is reference to "MPC7455 Rev2.0 (Apollo), 483 CBGA, 2.5V Bus, 2.5V L3".



    Among the menues in the

    <a href="http://e-www.motorola.com/cgi-bin/faq.cgi?kbase=powerpc"; target="_blank">FAQ</a> the 7445/7455 is called Apollo, and we see the codenames of other processors too.</strong><hr></blockquote>



    Try a search for Apollo <a href="http://e-www.motorola.com/collateral/SNDF2002RECAP_H1101.pdf"; target="_blank">HERE</a>



    page 17

    Goals of new G4 Technology Vehicle:

    The Apollo program

    ? Use proven SOI technology for higher performance design

    ? Use advanced 0.13µ technology (HiP7-SOI)



    What Motorola is calling Apollo now ain't there yet.
  • Reply 31 of 173
    programmerprogrammer Posts: 3,409member
    It looks to me like "Apollo" actually refers to all process improvements to the 745x design. The chips we have now are SOI, but not 0.13. There will likely be a 0.13 version showing up soon (Jan/Feb?).



    As for the 970, proper business practice says you don't lie to your clients and investors. IBM's best estimates are what they gave -- 1.8 GHz, ~1000 SPEC, shipping in volume 2H '03. They say these are estimates because the nature of the business always includes uncertainty, and their statements must reflect that. This doesn't change the fact that the best guess for the real numbers is what IBM's estimates are. They know way more about their new design and all the details of bringing it to market, so why does anybody here think they can possibly make a better estimate?! They specifically said that the SPEC estimates were conservative, so they might be slightly higher but I doubt it'll be by much.



    Based on this design's pipeline length and advanced process (with a fast migration to 0.09) and low power consumption I think we'll probably see its clock rate scale quite well. They don't need to make estimates that far in the future, so they're not going to speculate publicly. Considering, however, that the 7450 (which started at 733 MHz) will probably make it to 1.8 GHz with a 7 stage pipeline and 0.13 process, I'd guess that the 970 will start at 1.8 GHz and may see 3-4 GHz by the end of its life. After all, AMD is saying that Operton will make that level with a similar pipeline length and higher complexity level.
  • Reply 32 of 173
    eugeneeugene Posts: 8,254member
    [quote]Originally posted by rickag:

    <strong>



    Try a search for Apollo <a href="http://e-www.motorola.com/collateral/SNDF2002RECAP_H1101.pdf"; target="_blank">HERE</a>



    page 17

    Goals of new G4 Technology Vehicle:

    The Apollo program

    ? Use proven SOI technology for higher performance design

    ? Use advanced 0.13µ technology (HiP7-SOI)



    What Motorola is calling Apollo now ain't there yet.</strong><hr></blockquote>



    Sure it is. Just like the NASA Apollo program was designed to put a man on the moon, the Motorola program wasn't designed to complete every goal in one step. The first NASA Apollo missions merely tested the command module in orbit around Earth. They later tested docking with the lunar module...then they made passes around the moon, then Apollo 11 finally put a man on the moon. Later Apollo missions introduced further achievements like the lunar rover...



    The Motorola Apollo project is no different in this respect. The g4 doesn't need to be .13µ to be called Apollo, especially when the 745x/744x series are obviously known as Apollo chips internally.
  • Reply 33 of 173
    kidredkidred Posts: 2,402member
    [quote]Originally posted by Programmer:

    <strong>It looks to me like "Apollo" actually refers to all process improvements to the 745x design. The chips we have now are SOI, but not 0.13. There will likely be a 0.13 version showing up soon (Jan/Feb?).



    As for the 970, proper business practice says you don't lie to your clients and investors. IBM's best estimates are what they gave -- 1.8 GHz, ~1000 SPEC, shipping in volume 2H '03. They say these are estimates because the nature of the business always includes uncertainty, and their statements must reflect that. This doesn't change the fact that the best guess for the real numbers is what IBM's estimates are. They know way more about their new design and all the details of bringing it to market, so why does anybody here think they can possibly make a better estimate?! They specifically said that the SPEC estimates were conservative, so they might be slightly higher but I doubt it'll be by much.



    Based on this design's pipeline length and advanced process (with a fast migration to 0.09) and low power consumption I think we'll probably see its clock rate scale quite well. They don't need to make estimates that far in the future, so they're not going to speculate publicly. Considering, however, that the 7450 (which started at 733 MHz) will probably make it to 1.8 GHz with a 7 stage pipeline and 0.13 process, I'd guess that the 970 will start at 1.8 GHz and may see 3-4 GHz by the end of its life. After all, AMD is saying that Operton will make that level with a similar pipeline length and higher complexity level.</strong><hr></blockquote>





    From what I know the 970 will debut next fall and moto's new G4 will debut this Jan. We are at 1.25ghz, so we should see up to about 1.4ghz top from the G4. When IBM comes in with the 970 at announced 1.4ghz -1.8ghz speeds, that's inline with Apple dropping the top of the line to the bottom. So I would expect 1.4ghz, 1.6ghz and 1.8ghz 970 chips next fall. The G4 should be going to the iMac and PB leaving them plenty of updates.
  • Reply 34 of 173
    mokimoki Posts: 551member
    [quote]Originally posted by geekmeat:

    <strong> <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" /> lets talk a little bit about the new IBM 970 chip and the mythical motorola G5.

    I have bad news for all of you who were/are so excited by the announced IBM chip.

    first off all it isnt even coming out until late 2003 at the earliest.

    its initial clock speed will be 1.8 ghz.

    its specint rating is around 1000.

    </strong><hr></blockquote>





    ...which is roughly 3.5x better than the fastest current G4 processor (roughly 5.5x better for specfp). There is also some other cool news coming concerning this chip too... you shall see...



    [quote]Originally posted by geekmeat:

    <strong>i just want everyone to calm down a bit and THINK!

    this ibm chip IS NOT THE ANSWER.

    im sure im the only saying this but i feel in my gut that if motorola goes down....so does apple.

    my god bless us all and long live apple!!!!!</strong><hr></blockquote>



    The ibm chip is absolutely one answer -- and a very good answer at that.
  • Reply 35 of 173
    mokimoki Posts: 551member
    [quote]Originally posted by Matsu:

    <strong>IBM at least has its cards on the table</strong><hr></blockquote>



    There's another poker term that's worth bringing up: sandbagging



  • Reply 36 of 173
    [quote]Originally posted by moki:

    <strong>



    The ibm chip is absolutely one answer -- and a very good answer at that.</strong><hr></blockquote>



    "One answer?" So there's other answers?



    Moki, what are your thoughts about this theory that Motorola's next chip will fill the gap between now and Fall of 2003? I don't think I've seen you address that (yet) . . .
  • Reply 37 of 173
    mokimoki Posts: 551member
    [quote]Originally posted by CommonSense:

    <strong>Moki, what are your thoughts about this theory that Motorola's next chip will fill the gap between now and Fall of 2003? I don't think I've seen you address that (yet) . . .</strong><hr></blockquote>



    Honestly, I haven't been paying much attention to MOT, I wrote them off a while ago... if a "G5" comes out from MOT for Apple's desktop machines, I'll have a very hard time believing it.
  • Reply 38 of 173
    tjmtjm Posts: 367member
    [quote]Originally posted by moki:

    <strong>



    There's another poker term that's worth bringing up: sandbagging



    </strong><hr></blockquote>





    I had noticed on the last slide of the IBM presentation that there was an asterisk by the dates on the "Target Schedule", with the footnote "Estimation only, subject to change without notice". It seemed like grasping at straws to consider that it may come out before the targeted dates, so I didn't comment on it previously. Perhaps we'll get a surprise announcement of, "Gee, this really scaled up to production MUCH faster than we thought! (wink wink, nudge nudge)" sometime in the not-too-distant future (i.e. early next year)?
  • Reply 39 of 173
    [quote]Originally posted by Programmer:

    <strong>

    Based on this design's pipeline length and advanced process (with a fast migration to 0.09) and low power consumption I think we'll probably see its clock rate scale quite well. They don't need to make estimates that far in the future, so they're not going to speculate publicly. Considering, however, that the 7450 (which started at 733 MHz) will probably make it to 1.8 GHz with a 7 stage pipeline and 0.13 process, I'd guess that the 970 will start at 1.8 GHz and may see 3-4 GHz by the end of its life.</strong><hr></blockquote>



    Not sure where they got the numbers, but I've seen more than one <a href="http://zdnet.com.com/2100-1103-961862.html"; target="_blank">article</a> claiming the 970 will top out at 3 GHz, so maybe they are speculating.
  • Reply 40 of 173
    I don't think the sandbagging reference was to the production schedule, but rather that IBM is holding back on disclosing all of the chip's features and potential. Could be wrong though - it would be nice to buy this sooner than later.
Sign In or Register to comment.