Geekbench-results of the new iMacs (Mid 2010) compared to the old ones (Late 2009)

Discussion in 'iMac' started by Felias, Jul 29, 2010.

  1. macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #1
    Hi there,

    i know Geekbench only has a limited significance, but still is an interested value for comparison.

    When browsing through the Geekbench-results of the last 24 hours there are several entries of iMacs with the CPU-numbers of the new iMacs.

    From what i can see there average results are:

    NEW iMACs:

    32bit

    64bit:


    OLD iMACs

    32bit

    64bit

    How much performance do I get for my dollars? --> see here

    I'll try to keep the list updated, as more results are probably coming in today and in the next days to come.

    Please be aware: In Geekbench there always are some very strange readings, as people can give the name of the system themselves. The only relevant facts is the CPU used. Only the average number is relevant, which does not proof anything before there is a reasonable high number of results are available.
     
  2. macrumors regular

    Joined:
    Feb 28, 2008
    #2
    So you mean the difference between the i3 3.20 & i5 2.8 Quad is only 800 points?
     
  3. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #3
    There were about 4-5 entries for both the CPU-classes, but: yes, thats what the current results are saying. I find this strange as well, as i expected the new i5 to be closer to the old i7... lets just wait for more results maybe.

    Update: First result of a 2,93Ghz i7 added. About 500 points more than the average "old" i7. But this is just the first number, so it's just a hint, nothing more.

    Update2: All results are 32bit.
     
  4. macrumors member

    tturland

    Joined:
    Jan 23, 2007
    #4
    Have just searched on Geekbench and found a 600 point difference between the new i7 and a previous iMac i7....is it really worth the difference for a new i7 vs a refurb i7??
     
  5. macrumors 65816

    Joined:
    Jul 7, 2007
    #5
  6. macrumors regular

    Joined:
    Feb 28, 2008
    #6
    Hey, can you tell me why the difference in 32bit & 64bit benchmarks? I don't mean the real difference in points; just why there ARE 32bit & 64bit benchmarks. Aren't all these computers 64bit with Leopard and/or Snow Leopard?
     
  7. Moderator emeritus

    Hellhammer

    Joined:
    Dec 10, 2008
    Location:
    Finland
    #7
    32-bit cannot utilize more than ~3.3GB of RAM and AFAIK cannot take the full advantage of 64-bit CPU so we need 64-bit to get good result with no other apps open
     
  8. macrumors regular

    Joined:
    Feb 28, 2008
    #8
    And how do you 'set' it to 64bit? That's automatically i guess? Then why are there so many 32bit benchmarks?
     
  9. Moderator emeritus

    Hellhammer

    Joined:
    Dec 10, 2008
    Location:
    Finland
    #9
    You have to pay for the 64-bit version of GeekBench ;)
     
  10. macrumors regular

    Joined:
    Feb 28, 2008
  11. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #11
    -list updated-

    I don't know how the guys from Geekbench summarized the numbers for the late 2009-model, but it doesn't match the readings i can see in the results browser... i've put my own readings in there instead.

    Update: Okay, screw my own list, i've linked the results from the browser. I haven't found a way to differ between 32bit and 64bit in the search, so both are combined.
     
  12. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #12
    I guess we can say now that the 32bit-results are indeed poor... but when running 64bit (which will be the situation), the difference is a lot higher (1400 points)
     
  13. macrumors G4

    Joined:
    Mar 4, 2006
    #13
    I take it nobody has yet received a 21.5" i5 machine then. Impressive improvements all the same. Certainly knocks on the head this idea that the i3 offers no extra performance over a C2D.
     
  14. macrumors regular

    Joined:
    Mar 3, 2009
    #14
    At first glance, the i3 results are certainly impressive compared to the previous generation. Although the performance difference between the 3.06 and 3.2 processors does seem a bit marginal. Not sure it justifies the increased asking price if all you're after is speed...
     
  15. macrumors newbie

    Joined:
    Jul 28, 2010
    #15
    Hmm, so how are these scored compared to other computers then mac?
    And what seems to be the best buy here?
     
  16. macrumors 6502

    Joined:
    Dec 7, 2007
    #16
    How is there iMac i7-870 in that list tested 8 months ago? They only came out last week.
     
  17. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #17
    Probably someone created a hackintosh and put in an i7-870 and called it an "iMac"... as the name can be set freely. Maybe even a Windows-System that has been named "iMac" by the user.
     
  18. macrumors 6502

    Joined:
    Dec 7, 2007
    #18
    Ok thats what I figured.
     
  19. macrumors member

    Joined:
    Jun 30, 2010
    #19
    lets just say on average the refresh i7 has 800 better benchmark than pre-refresh i7 is that a big different i dont know much about benchmark stuff? because i really debating if not i should try get upgraded to the refresh version or stick with my current i7 2.8
     
  20. macrumors 6502

    Joined:
    Aug 1, 2004
    Location:
    Nara, Japan
    #20
    Do not place too much emphasis on 400 or even 600 points difference in Geekbench. I have a Dual G5 that scored 2500 and an old iMac Core 2 Duo that scored 2900 points and I'll swear that the G5 seemed faster to me. You probably need a difference of 700 or even a 1,000 points before you can see any real difference with your eyes and even then it's probably got to be processor or memory intensive process.
     
  21. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #21
    The benchmark is just a hint.

    If you have a flawless iMac and/or you're out of the time to give it back, i would keep it. All the current issues with the iMacs (noise, yellow-tint, gray banding,...) are more important than a 5-10% speed bump.
     
  22. macrumors regular

    Joined:
    Jul 5, 2010
    Location:
    New York
    #22
    So back to the ultimate question for people interested in the 21.5" models. Is the i5 3.6 worth it, or is it better to stick to the i3 3.2.

    FYI, if you have an educational discount, the apple store discounts the i5 upgrade from $200 to $180. Not a whole lot but appreciated.
     
  23. thread starter macrumors member

    Joined:
    Jul 29, 2010
    Location:
    Germany
    #23
    buy it, benchmark it, and let us know :) Then we'll tell you.

    Or wait until someone else does...
     
  24. macrumors G4

    Joined:
    Mar 4, 2006
    #24
    Someone in one of the other threads was waiting for theirs to be delivered today. A bit of gentle persuasion by bombarding them with PMs linking to Geekbench might get them to post some results :)
     
  25. macrumors 65816

    Joined:
    Jan 29, 2008
    Location:
    Chicago
    #25
    Hmmm so new i5 versus old i5 is about an 8% to 9% difference in ideal conditions. Not worth paying an extra $470 for, IMO, but it is a good bump.

    I got a feeling the next Mac revision will be early to mid 2011 and contain a similar small bump, followed by a major revision late 2011 or early 2012.
     

Share This Page