Unreal Tounament 2004, Doom3 Universal Benchmarks: Core Duo ...

Discussion in 'MacBytes.com News Discussion' started by MacBytes, Jan 31, 2006.

  1. MacBytes macrumors bot

    Jul 5, 2003
  2. Fiveos22 macrumors 65816


    Nov 20, 2003
    Wow. Just...wow.

    Now if only I could learn to love UT2k4...UT original still has my heart.
  3. greatdevourer macrumors 68000

    Aug 5, 2005
    Maxed out and 1024x768 and still getting over 50fps? Sounds like I'm set for mobile gaming fun :D What I really want to see is someone test Halo in emulation (in the highly unlikely scenario that they release a Unibin, it'll probs be ages)

    EDIT: I just saw the Halo specs on barefeats. Ouch :(
  4. socamx macrumors 6502

    Oct 7, 2004
    I don't want to sound negative...but why is everyone impressed so much with benchmarks like this? The new iMacs have a serious advantage, a second core. (I'm directing this to the people who gloat about the new Intels being that much better than PPC.)

    I'd like to see benchmarks with the second core disabled for a more fair comparison to the G5 iMacs, and this goes with anything benchmark wise.

    Doesn't surprise me that a computer with a second core performs that much better...
    Personally I think any of the benchmarks that show the new Intel iMacs doing better has more to do with the fact it is dual core than the fact it is the Intel.

    As a side note...the graphics card in the new iMacs are also significantly better than the G5's. So that also has to be accounting for something.

    Either way though, performance gains are always nice.
  5. MacBandit macrumors 604


    Aug 9, 2002
    Springfield, OR (Home of the Simpsons)
    The gains are actually more likely from the much much improved graphics processor as opposed to the second processor which does very little to help Unreal Tournament.
  6. Nermal Moderator


    Staff Member

    Dec 7, 2002
    New Zealand
    Turning the second core off is *not* a fair comparison. What's the point in comparing against a configuration that nobody will use in the real world? People are going to run their systems with both cores enabled, so that's what we need to compare against.
  7. Fiveos22 macrumors 65816


    Nov 20, 2003
    Is that so? Well then it might surprise you that the Power macs were in dual processor setups when UT2k3 came out (then 2k4). They didn't get nearly this good of a frame rate...(I know, I've tried). The port from x86 to PPC is mediocre to poor for most games. I've seen games play beautifully on PC hardware that is roughly equivalent to my mac setup where they typically look modest.

    A 40% improvement in performance fps-wise is astounding. Ask most mac gamers (on their powerbooks, ibooks, and imacs) what kind of fps they get, and they say 24-30 on normal settings. (One obvious exception being Quake 3, however that is kinda long in the tooth at this point).

    Furthermore UT2k4 (and I believe 2k3) were built so that the sound component (which is a huge cpu process) could be dedicated to a second processor (or a second thread-taker, in the case of hyper-treading). Assessing a dual core cpu in a game is not cheating, its assessing a dual core cpu. And in this case, the point of the article is to assess how ostensibly the same program can perform when it is optomized properly for the hardware.
  8. bousozoku Moderator emeritus

    Jun 25, 2002
    Gone but not forgotten.
    Since x86 machines didn't usually come in dual processor configurations UT2003/UT2004 wasn't designed to really take advantage of such a thing and the port didn't do a major re-design to improve performance with dual processors. They should have put each bot on its own thread--that would have been great with the quad PowerMac--but it wasn't to be.

    Now that dual core processors are making inroads in the x86 world, we should see better game designs to take advantage of them.

    It's great to see that preliminary testing shows such a huge leap. I think we'll all be pleased to see some parity with the Wintel crowd, as far as performance goes.
  9. socamx macrumors 6502

    Oct 7, 2004
    I know that, but there are people out there saying that the Intel chips absolutely spank the PPC chips. For those people I direct that to. Compare single to single before saying the Intels are that much greater.
  10. nagromme macrumors G5


    May 2, 2002
    "Overall the new 2.0GHz Intel iMacs provide a significant speed bump with the universal binary for Unreal Tournament 2004 (UT2004), providing a huge 40%+ benefit over the 2.1Ghz PPC G5 in the botmatch benchmark."

    Count me in! :)

    Yes, the new iMacs have two CPU cores and a faster GPU. Those are GOOD things and perfectly fair to compare to. Nobody's saying the G5 is a bad chip--it's not. But you can't get two in an iMac and you can't get any in a laptop.

    Anandtech did extensive testing of G5 vs. Core Duo, and results varied a lot by app (even Universal apps), but with one core disabled, the Core "Solo" tended to be slower than a G5--by a small amount. However, there are TWO cores. For the same price we used to pay for one. So it's fair to test that way.

    (PS: where are these matinee fly-bys found?)
  11. SiliconAddict macrumors 603


    Jun 19, 2003
    Chicago, IL
    Mmmm MacBook Pro gaming goodness. Its a mouthful to say but what a tasty one. *counts the minutes til his MBP arrives* :D
  12. nospleen macrumors 68000


    Dec 8, 2002
    I can understand what you are saying, but Nermal was dead on. But, what are the dual core G5's benchmarks running these games? I guess this would give us a good comaprison. Does anyone know?

Share This Page