IBM's Latest Fab!

2»

Comments

  • Reply 21 of 34
    backtomacbacktomac Posts: 4,579member
    [(and as possible leverage against Intel should it ever be necessary).



    The leverage against intel is AMD
  • Reply 22 of 34
    Quote:

    Originally posted by OverToasty

    There ya have it folks - a definite "maybe" ... nothing wrong with that, so ...



    Apple would be smart (in my very humble opinion) to keep its options open with this Univeral Binary thing for a while, maybe even experiment with tweaking for the Cell Model and see what happens: and yes, that's NOT just for backwards compatability.




    If you think that supporting Cell is just some slight change of work on the dual binaries, then dream on. Many Apps have to be rewritten to utilize Cell's performance. Because Cell itself has a very different structure to other older PowerPCs.



    Besides, you are talking about keeping three different technologies compatible. How about the new infinite adding features for each technology in the future? Are keeping these features from each tech compatible easy for you?



    If you make it, I will like to hear good news from you.
  • Reply 23 of 34
    Quote:

    Originally posted by spliff monkey

    MY point was obviously missed. Many have told me and others on these boards (somewhat overconfidently) that you's never see cell in a desktop and or server. Well here they are! Any reason apple couln't or wouldn't make this work? Please I just want to know whay my next "PowerMac" won't be using this technology, rendering entire cities in real time. OMFG how could that not be appealing. Obviously I'm ignorant to the tecchnical/ political reasons why not but I'm not dumb enough to not recognize that it is very obviously possible. I just want to understand are intel chips superbly better or are we in the middle of another snow job?



    If you need a Cell desperately, just go for PS3.



    I don't see why Apple must have the techs you want in their computer. Is it an obligation for Apple to put the most new advanced techs in their product line?



    Are Intel chips superbly better? I don't think that's a good question. How do you define "better" anyway? At least I don't assume Apple would use "bad" or "worse" components in their computer.



    I think Apple could continue produce great products. That's all I want.
  • Reply 24 of 34
    backtomacbacktomac Posts: 4,579member
    [QUOTE]Originally posted by mechengit

    [B]



    Are Intel chips superbly better? I don't think that's a good question. How do you define "better" anyway? At least I don't assume Apple would use "bad" or "worse" components in their computer.



    Your reply is 'Clinton' like and avoids the question-would the cell be faster or cooler running chip at the same level of performance. I'm not convinced by what i've read at forums but the question is legitamate for debate.
  • Reply 25 of 34
    [QUOTE]Originally posted by backtomac

    [B]
    Quote:

    Originally posted by mechengit





    Are Intel chips superbly better? I don't think that's a good question. How do you define "better" anyway? At least I don't assume Apple would use "bad" or "worse" components in their computer.



    Your reply is 'Clinton' like and avoids the question-would the cell be faster or cooler running chip at the same level of performance. I'm not convinced by what i've read at forums but the question is legitamate for debate.




    I didn't tempt to avoid any questions. I don't see how good you are at mind reading. All I want to know is what means "better" in comparsion of x86 chips and Cell.



    Besides, I don't even care about the question you mentioned. So what? Does that mean every computer companies should put Cell into their product line?
  • Reply 26 of 34
    backtomacbacktomac Posts: 4,579member
    Quote:

    Originally posted by mechengit

    I didn't tempt to avoid any questions. I don't see how good you are at mind reading. All I want to know is what means "better" in comparsion of x86 chips and Cell.



    Besides, I don't even care about the question you mentioned. So what? Does that mean every computer companies should put Cell into their product line?






    I'm not trying to read any minds. The question asked was "is the cell better" than x86. It's obvious that better refers to performance or speed. Why not come out and say yeah or nay on the issue? Saying you don't know what the person means by better reminds me of a former president who wanted to know "what the definition of is, is."



    I said that I'm not convinced that cell is faster than x86 for general cpu use based on what I've read on forums. I don't think Apple should use them either. Most users wouldn't be swayed to Apple computers with RISC chips powering them even if they were faster than comparably priced pcs with x86 chips. History has shown that even when the ppc had a performance advantage over it's x86 counterpart it really didn't translate into increased market share for Macs.
  • Reply 27 of 34
    Quote:

    Originally posted by backtomac

    I'm not trying to read any minds. The question asked was "is the cell better" than x86. It's obvious that better refers to performance or speed. Why not come out and say yeah or nay on the issue? Saying you don't know what the person means by better reminds me of a former president who wanted to know "what the definition of is, is."



    But you miss a point, even performance or speed of two different structures could not be compared easily.

    Anyone(LIKE Sony or Apple) could play many fancy numbers while hiding many facts.



    You think the "better performance" is very easy to define? To me it's not "what the definition of is, is." Many assumptions have to be defined before you can compare two different structures.



    Still, what do you mean by "better"?

    The word "better" is still meaningless even if you are obviously talking about performance or speed.



    Unit/Time? What does "Time" include? What does "Unit" include? There are lots of assumptions hiding between some simple words or numbers. Also, have software developers of a single software program truly utilized each CPU's performance? If not, how do you compare performance between these two different structures?



    I can go on forever...
  • Reply 28 of 34
    thttht Posts: 5,421member
    That's not IBM's latest fab. That's IBM's fab in 2007. It's pretty safe to say that IBM's fab isn't as good as AMD (who they share the technology with), let alone Intel.



    It's nice that they can fab chips with SOI and strained silicon at 65 nm, but the proof is in the pudding. Apple already is shipping 65 nm mobile dual-core processors in their machines. They don't be needing to wait a year for them.



    If Power6 is as good as they say it is, great for them, but that processor is in an entirely different market than Apple is. As far as a "G6" processor is concerned, everything I've seen from the PPE indicates that Intel will still have better performance per watt, plus the all important advantage of shipping 65 nm and 45 nm CPUs much earlier than anyone else.
  • Reply 29 of 34
    backtomacbacktomac Posts: 4,579member
    Quote:

    Originally posted by mechengit

    But you miss a point, even performance or speed of two different structures could not be compared easily.

    Anyone(LIKE Sony or Apple) could play many fancy numbers while hiding many facts.



    You think the "better performance" is very easy to define? To me it's not "what the definition of is, is." Many assumptions have to be defined before you can compare two different structures.



    Still, what do you mean by "better"?

    The word "better" is still meaningless even if you are obviously talking about performance or speed.



    Unit/Time? What does "Time" include? What does "Unit" include? There are lots of assumptions hiding between some simple words or numbers. Also, have software developers of a single software program truly utilized each CPU's performance? If not, how do you compare performance between these two different structures?



    I can go on forever...








    Your point that it is difficult to compare different chips that run under different structures is a valid point. But it is still possible to test the chips and find their strengths and weaknesses. While the tests may not be a slam dunk victory for any particular chip the results are generally beneficial.
  • Reply 30 of 34
    onlookeronlooker Posts: 5,252member
    Quote:

    Originally posted by THT

    That's not IBM's latest fab. That's IBM's fab in 2007. It's pretty safe to say that IBM's fab isn't as good as AMD (who they share the technology with), let alone Intel.



    It's nice that they can fab chips with SOI and strained silicon at 65 nm, but the proof is in the pudding. Apple already is shipping 65 nm mobile dual-core processors in their machines. They don't be needing to wait a year for them.



    If Power6 is as good as they say it is, great for them, but that processor is in an entirely different market than Apple is. As far as a "G6" processor is concerned, everything I've seen from the PPE indicates that Intel will still have better performance per watt, plus the all important advantage of shipping 65 nm and 45 nm CPUs much earlier than anyone else.






    And on top of that there is no reason for IBM to make a 980, or whatever the G6 would be called. They are no longer making processors for Apple, and no one else is big enough for them to make one for that uses this particular processor design. Power6 is what it is, but the Power4, or Power5 Processors have never made their way into a Mac yet, and neither will this one.
  • Reply 31 of 34
    rickagrickag Posts: 1,626member
    Quote:

    Originally posted by Programmer

    I would be careful with those characterizations. Cell will actually do just fine at many computing tasks which are considered "general purpose"; the key is that "strange programming model". The important observation should be that unmodified software will generally not run particularly well on a Cell, although a 4+ GHz in-order processor can do pretty well compared to an ~2 GHz out-of-order one, especially given the >25 GB/sec memory bandwidth of the Cell. If the software is redesigned/rewritten to conform to the Cell's "strange programming model" then for many tasks (floating point or integer) where people care about performance, the Cell will perform extremely impressively. And only time will tell how long the Cell's programming model will remain strange... the hardware design pressures driving it are not unique to STI.



    Programmer, I respect your knowledge and have learned much from you. In the future it seems Cell's model will be more and more accepted. How long this takes, I have no clue. And I feel that Intel will be able to keep pace without too much trouble.



    But, as far as Apple's concerned, I will not worry too much about IBM's Cell processor until Microsoft ports its' operating system over to it.
  • Reply 32 of 34
    snoopysnoopy Posts: 1,901member
    Quote:

    Originally posted by onlooker

    And on top of that there is no reason for IBM to make a 980, or whatever the G6 would be called. They are no longer making processors for Apple, and no one else is big enough for them to make one for that uses this particular processor design. . .





    IBM might make one for their blade servers, if it catches on. I don't know how well blade servers with the 970 have sold, or how well the blades with the 970MP will be selling. My impression is that the 970MP blades use fewer chips than Apple buys for the Power Mac.



    What will they call the next generation Power Macs? The Pro Mac? Just kidding.



    Jerry
  • Reply 33 of 34
    Ya'll are crazy! mechengit your really crazy man! You take things way to seriously (and that's coming from a very serious guy) but thanks for all of the info! Sincerely! I have a much better understanding of the scenario now, cell, power 6, G6 etc. still seems a viable alternative, in whatever variant it would incarnate itself, but it still depends where the money is being invested. I guess that's the reason for my question, why are computer users always being "held back?" (Of course were not but in spirit we truly are, especially when you consider that computers perhaps made things faster but now were doing twenty things at once; bottle neck! I hate waiting for computers (somehow it seams so much fair to wait for sentiant beings, I think because it's actually polite but why bother extending the coutesy to a beige box. Frankly I wish their presence would disappear from my user level awareness, too often ion my everyday do they remind me they are there, fine when I need to cruch large data sets for a virtual sound stage but not when I want to pay a bill) Ultimately it's all about the money, who's invested into what, how committed are they, how much work has to be done to get the tech working properly. I'm an extreme perfectionist, when I see something not working quite right I want to fix it, the latest Mac OS is less than perfect but it's much more stable and capable than OS9 even if it's clunky in it's UI, I would expect similar trade off with the intel transition, not as robust, quicker for the user overall but not a big number cruncher. One of these days just for kicks I'd like to see someone make a computer and OS that really trashed the old model and broke all the rules! that's what Apple did back in when they were a fearless startup but I guess I have to get over the fact that they will never be able to do that again, because they're invested now and to abandon all that they have in the hopes of developing something great is far to risky for anyone except those that have nothing to lose, much like the Apple from 1984.
  • Reply 34 of 34
    Quote:

    Originally posted by snoopy





    What will they call the next generation Power Macs? The Pro Mac? Just kidding.



    Jerry [/B]





    I hope so "PROMAC" the happy mac that convinces you it's actually going to make your life fulfilled!
Sign In or Register to comment.