The G4.

dartblazerdartblazer
Posted:
in Future Apple Hardware edited January 2014
[code]

1 \tOctober 1999\t G4 350

5\t February 2000\t G4 500

10\t July 2000\t\t G4 500 MP

16\t January 2001\t G4 533 MP

17\t February 2001\t G4 733

22 \tJuly 2001 \t\t G4 800 MP

29\t February 2002\t G4 1000 MP

</pre><hr></blockquote>







Next G4 update in 4 to 5 months? They've been consistent, why change now.



[ 03-15-2002: Message edited by: dartblazer ]</p>
«1

Comments

  • Reply 1 of 21
    quaremquarem Posts: 254member
    [code]

    1 \tOctober 1999\t G4 350

    5\t February 2000\t G4 500

    10\t July 2000\t\t G4 500 MP

    16\t January 2001\t G4 533 MP

    17\t February 2001\t G4 733

    22 \tJuly 2001 \t\t G4 800 MP

    29\t February 2002\t G4 1000 MP

    </pre><hr></blockquote>







    Next G4 update in 4 to 5 months? They've been consistent, why change now.



    [ 03-15-2002: Message edited by: dartblazer ][/qb]&lt;hr&gt;&lt;/blockquote&gt;



    You should either quote the top of the line or the bottom of the line always. You have the 350, while it should have been the 450, and just leave the 533 out of it, since it was BTO and the 733 was the top of the line.



    [ 03-15-2002: Message edited by: Quarem ]</p>
  • Reply 2 of 21
    thuh freakthuh freak Posts: 2,664member
    if its too predictable, then they have no surprises. no surprises is bad for business, b/c people won't buy comps at all during the last few weeks, while awaiting the new revision. i heard last year that Apple was goign to do a lot more between expo announcements, when people are expecting them; so they can keep some allure. people were (are) too focused on the expos, and their seemingly patterned release schedule.



    -"George Dietrich?"

    Yes, that's me.
  • Reply 3 of 21
    junkyard dawgjunkyard dawg Posts: 2,801member
    Wow that is one butt-ugly graph! Did you make that in Excel or something? Ugh. That's why I hate Excel for graphing, it's aimed at business dorks and not scientists or engineers. Business tools like to impress people with 3d graphs and other fantastical bullsh!t that obscures the data and looks ugly.



    Ahem, and what on earth are "series 1", and "series 2"? Always label your graphs, unless you want to trick people, and then it's best to leave them unlabeled and without a legend.



    It would also be interesting to see the graphs for both AMD and Intel. You could compare the slope of each graph, and then plot slope vs. R&D budget to see how strong the correlation is.



    Even better, would be to find performance benchmark data for each of the G4 iterations, and plot both the benchmark data and the MHz data on the same graph. Actually that would be VERY interesting. I may even do that if I'm inspired and not too drunk tonight.
  • Reply 4 of 21
    keilwerthrebornkeilwerthreborn Posts: 129member
    [quote]Originally posted by Junkyard Dawg:

    <strong>Wow that is one butt-ugly graph! Did you make that in Excel or something? Ugh. That's why I hate Excel for graphing, it's aimed at business dorks and not scientists or engineers. Business tools like to impress people with 3d graphs and other fantastical bullsh!t that obscures the data and looks ugly.



    Ahem, and what on earth are "series 1", and "series 2"? Always label your graphs, unless you want to trick people, and then it's best to leave them unlabeled and without a legend.



    It would also be interesting to see the graphs for both AMD and Intel. You could compare the slope of each graph, and then plot slope vs. R&D budget to see how strong the correlation is.



    Even better, would be to find performance benchmark data for each of the G4 iterations, and plot both the benchmark data and the MHz data on the same graph. Actually that would be VERY interesting. I may even do that if I'm inspired and not too drunk tonight. </strong><hr></blockquote>



    What an extraordinarily productive post!
  • Reply 5 of 21
    eugeneeugene Posts: 8,254member
    Your graph is on some serious crack.



    1) The G4 debuted at 450 MHz in August 1999 if you don't count the 500 MHz debacle.



    2) Your graph is totally off in terms of timescale. You have 533 MHz and 733 MHz separated by a month, but you have the gap between them the same as any other gap.



    3) Why even separate the 533 and 733 at all, the announcements were made at the same time...



    And what is that second curve?
  • Reply 6 of 21
    undotwaundotwa Posts: 97member
    [quote]Originally posted by Junkyard Dawg:

    <strong>Wow that is one butt-ugly graph! Did you make that in Excel or something? Ugh. That's why I hate Excel for graphing, it's aimed at business dorks and not scientists or engineers. Business tools like to impress people with 3d graphs and other fantastical bullsh!t that obscures the data and looks ugly.

    </strong><hr></blockquote>



    I think that's AppleWorks.
  • Reply 7 of 21
    eugeneeugene Posts: 8,254member




    Accurate to within a month. I chose to put the initial announcement at 500 MHz because that's what we were originally promised. I used announcement dates instead of ship dates.



    It's ultra blurry because I was lazy and scaled down a large graph in haste.



    [ 03-16-2002: Message edited by: Eugene ]</p>
  • Reply 8 of 21
    brussellbrussell Posts: 9,812member
    Now I'd like to see Intel's Mhz in the same time frame.
  • Reply 9 of 21
    bjnybjny Posts: 191member
    [QUOTE]Originally posted by Eugene:

    [QB]Your graph is on some serious crack.



    1) The G4 debuted at 450 MHz in August 1999 if you don't count the 500 MHz debacle.



    G4 began with 350 MHz (remember "Yikes" motherboard), next was 400MHz as so many of us own awaiting the G5
  • Reply 10 of 21
    eskimoeskimo Posts: 474member
    [quote]Originally posted by BRussell:

    <strong>Now I'd like to see Intel's Mhz in the same time frame.</strong><hr></blockquote>



    <a href="http://www.rit.edu/~mcm5276/cpuscale.htm"; target="_blank">here you go</a> BRusssel. I added AMD to the mix just for fun. I used the press releases from the various companies along with my own knowledge of events. For the G4 I just used Eugene's data. Note on the Intel chart even though they "announced" their 1GHz processor in March of 2000 they didn't actually ship it until August. Also I left off the 1.13GHz P3 as it was shortly recalled.



    Edit: someone told me Mac IE doesn't display my page right (doesn't show legend at bottom). The red line is the G4, the blue is Intel, and the green is AMD.



    [ 03-16-2002: Message edited by: Eskimo ]</p>
  • Reply 11 of 21
    brussellbrussell Posts: 9,812member
    Interesting. From when the G4 finally started getting past 500, the slopes are roughly equivalent.



    It's that flat spot at 500 that really killed the G4.
  • Reply 12 of 21
    serranoserrano Posts: 1,806member
    [quote]Originally posted by BRussell:

    <strong>Interesting. From when the G4 finally started getting past 500, the slopes are roughly equivalent.



    It's that flat spot at 500 that really killed the G4.</strong><hr></blockquote>



    never would've known that without the graph
  • Reply 13 of 21
    brussellbrussell Posts: 9,812member
    [quote]Originally posted by janitor:

    <strong>never would've known that without the graph </strong><hr></blockquote>

    Yes, but whenever a pretty colored graph tells me what I already new, it just gives me a warm feeling inside.
  • Reply 14 of 21
    moogsmoogs Posts: 4,296member
    You guys are still getting the G4/500 timeline wrong. Unless you count the screw up where they were announced and then pulled, this variant wasn't available (and didn't stay available) until February of 2000. So 02/2000 should be your "start date"....
  • Reply 15 of 21
    trumptmantrumptman Posts: 16,464member
    What is amazing is that I remember when Apple was getting so far behind the pack before the started shipping most machines with duel processors standard. Thus when Intel and AMD were at 1 gig, we had duel 500's and duel 450's giving us the equivelent of 900 and 1000 mhz respectively.



    Since then we are clearly behind again and Apple only offers duels with the premium system. Honestly there should be duels across the line. Apple is likely using the cost of the duels to keep up their insanely high 30% margins and hurt the marketshare when these machines don't move. Duel 933's and duel 800's would put a cork in many of the complaints about the Powermac line. It would make them 1600, 1866 and 2000 mhz respectively and that still allows plenty of room for product tiers.



    Apple should also already be updating the iMac to 1000, 900 and 800 low end. Apple apologists will say that the machines are just now starting to fill the channel. However the fact that Apple takes 3 months to ship a machine doesn't excuse the lack of updates. In their previous financial report Apple stated that they wanted to get away from the huge events every six months for updates because it caused machines to stack up before each of the trade shows. They wanted more low key events like the iBook introduction.



    There should be low key iMac update and MWTK should have Powerbook and iBook upgrades pronto.



    Nick
  • Reply 16 of 21
    cliveclive Posts: 720member
    To get this right I think you should take into account that the G4 didn't ship at 500 until into 2000 some time (though was due to ship by October 1999 if i remember correctly).



    You should also take into account the dual processors, separating them into a separate line - so the fastest single processor is currently 933 and the fastest ever dual 2*1GHz.



    I suppose we should do a comparison to established benchmarks too?
  • Reply 17 of 21
    eskimoeskimo Posts: 474member
    [quote]Originally posted by Clive:

    <strong>To get this right I think you should take into account that the G4 didn't ship at 500 until into 2000 some time (though was due to ship by October 1999 if i remember correctly).



    You should also take into account the dual processors, separating them into a separate line - so the fastest single processor is currently 933 and the fastest ever dual 2*1GHz.



    I suppose we should do a comparison to established benchmarks too?</strong><hr></blockquote>



    I changed the G4 line to reflect that it was initially introduced at 450MHz. However I think it wouldn't be worthwhile to plot the dual processor MHz since the Intel and as of last summer AMD solutions were also available in dual configurations. Double all the numbers if it makes you feel better.
  • Reply 18 of 21
    emaneman Posts: 7,204member
    [quote]Originally posted by BJNY:

    <strong>



    G4 began with 350 MHz (remember "Yikes" motherboard), next was 400MHz as so many of us own awaiting the G5</strong><hr></blockquote>



    Originally the G4 was to begin with a 400 Yikes model, then the whole 500MHz problem came about and that's when the 350 was introduced.
  • Reply 19 of 21
    eugeneeugene Posts: 8,254member
    The 450 MHz G4 was available almost immediately. My friend bought one the week of the announcement. He's lucky because they clocked the models down 50 MHz shortly afterward.



    For all purposes, you could treat Apple's G4 announcement as September announcement anyway. And the G4/450 shipped in small quantities before October.



    I don't care when the 500 MHz actually started shipping. Motorola was producing these chips (with errata) at announcement on August 31, 1999. Apple was set to release it in November of 1999. It's better to go by announcement dates rather than shipping dates.
  • Reply 20 of 21
    failedmathematicianfailedmathematician Posts: 95member
    I thought I would put together a graph of mathematica benchmarks, but I couldn't really gather enough info. But check out this web site:



    <a href="http://fampm201.tu-graz.ac.at/karl/mma.html"; target="_blank">mathematica timings</a>.



    The decline from total domination in Mathematica 2, to highly competitive in 3, to just flat out humiliation in 4 is obvious. Sometimes you don't need a pretty chart.
Sign In or Register to comment.