Why we might win the speed race yet

blackcatblackcat
Posted:
in Future Apple Hardware edited January 2014
Those of us who are optimists truelly believe a 1.5Ghz G4 on DDR will flatten the best x86 can offer, because it's about more than Mhz.



If you read <a href="http://www.theinquirer.org/?article=4423"; target="_blank">this Inquirer article</a> you'll find we might be closer than we thought to proving our point.



I doubt X would fall over at 2.8Ghz+

Comments

  • Reply 1 of 11
    bombthroatbombthroat Posts: 35member
    I don't think we'll see a problem like this with Apple hardware or software. Considering Apple makes BOTH parts (sells the hardware and software), it's unlikely they'd let a glaring mistake like this slip past them. Not to mention the article is talking about Windoze, please, we have OS X (UNIX) not some piece of crap from Micro$oft.
  • Reply 2 of 11
    blackcatblackcat Posts: 697member
    It's interesting to note that Win9x had exactly this problem a few years back when AMD released the K5. It seems M$ used for(; loops for timing, which is not good coding practice.



    So we'll hit 2.8Ghz sometime in the future and X will keep on going



    I wonder if dual or quad G4s would cause problems?
  • Reply 3 of 11
    rhumgodrhumgod Posts: 1,289member
    I remember while working for a software company that a piece of software we acquired in a buyout of another company would crash on CPUs &gt; 266MHZ (P-IIs back in those days). The software was written to receive mathematical computations, and if the computation wasn't completed in time, divide by zero errors, and such were popping up all over the place on newer sytems. Badly written software was the culprit, and it wouldn't surprise me in the least that this is a result of the same thing. Come on, did we all think that Windows was a well written piece of code, after all? It's just that technology is showing the weakness in their code. Laughable <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />
  • Reply 4 of 11
    gardenofearthlydelightsgardenofearthlydelights Posts: 963member
    So what happens when Intel introduces its <a href="http://www.theregister.co.uk/content/3/26194.html"; target="_blank">3.0GHz Pentium 4</a>? Will it bring Windows to its knees?



    Nahh... it'll probably just make AI posters cry for the mythical G5.
  • Reply 5 of 11
    nonsuchnonsuch Posts: 293member
    "My computer is so fast, my OS won't even run on it!" How's that for bragging rights?
  • Reply 6 of 11
    rhumgodrhumgod Posts: 1,289member
    [quote]Originally posted by Nonsuch:

    <strong>"My computer is so fast, my OS won't even run on it!"

    </strong><hr></blockquote>



    <img src="graemlins/embarrassed.gif" border="0" alt="[Embarrassed]" /> I guess it's all back to DOS for the Wintel crowd from here on out!
  • Reply 7 of 11
    spartspart Posts: 2,060member
    Wouldn't it be cool if it took MS years to work this out?



    Ahh, I can dream at least.



    <img src="graemlins/smokin.gif" border="0" alt="[Chilling]" />



    [EDIT: This is my 666th post. Does that mean that I'm evil?]



    [ 07-15-2002: Message edited by: Spart ]</p>
  • Reply 8 of 11
    nitridenitride Posts: 100member
    I think the goal for OS X is for it to do as little as possible on the CPU in any given time.



    The upshot is that bursts of speed are readily available when you start to encode an MP3 or durn a DVD.



    OS X itself should scale very well with CPU speed while various random apps may choke. The for() loop counting situation happened to Frontier I believe running on early G3s. Essentially they overflowed a short in a loop used to time out the splash screen.



    Carbon now has timers and auto-dismissing dialogs so that particular problem is easily avoidable.



    Computational quirks related to massive speeds may require special handling for situations where functions return before the rest of the code expects them to.



    Deferred processing and/or returning data is already a problem with a trully preemptive OS such as OS X, so its not like it will be a totally new problem for Mac OS X coders on really fast future CPUs.
  • Reply 9 of 11
    macommentarymacommentary Posts: 196member
    The other problem appears t obe heat and the PPC processors and apple have never had problems running on low power conumption or on keeping cool. Of course this is in relation to the Intel and AMD chips. Heck, when the P4 burst into the market needing a larger case to fit is heat sinks and fans alone, apple had just been through a line of computers (cube) that had no fan at all and could fit in about 1/4 of the P4 wintel box. heat shouldn't be an issue either. Power consumption not being a large issue either, it looks like we'll dodge this bullet completely
  • Reply 10 of 11
    yurin8oryurin8or Posts: 120member
    [quote]Originally posted by Blackcat:

    <strong>

    I doubt X would fall over at 2.8Ghz+</strong><hr></blockquote>



    Im posting this on a P4 2.26B overclocked to 2.9jiggamahertz on XP...and it's stable.



    [Edit: OMG I just BSOD'd. Better bump it back to 2.6ghz]



    [Edit: LOSERS ]



    [ 07-16-2002: Message edited by: yurin8or ]</p>
  • Reply 11 of 11
    bartobarto Posts: 2,246member
    Thats one of the funniest things I've seen today. The article and yurin8or.



    "Pentium 4 2.8GHz



    NOT Designed for Microsoft Windows"



    <img src="graemlins/lol.gif" border="0" alt="[Laughing]" />



    Barto
Sign In or Register to comment.