MorphOS 3.0 Quake III benchmarks
  • Yokemate of Keyboards
    Yokemate of Keyboards
    amigadave
    Posts: 2795 from 2006/3/21
    From: Northern Calif...
    I am not sure I understand the questions being asked in this thread, or the surprise that the video driver in MorphOS3.0 is not as fast as some demostration of an unfinished driver some time in the past.

    It was clear to me that the release of MorphOS3.0 would not include the updated video driver(s) and that 3D support for the PowerBook's Radeon 9700 Mobility video card, as well as 3D support for one or two other currently 2D only supported Radeon video cards, would come in a later release.

    I guess I must have missed the posting, or announcement from any of the MorphOS Dev. Team members regarding better video card driver performance that was supposed to give us this 140fps performance in Quake III.

    Of course I am not much of a game player, so I am not disappointed by the performance of MorphOS3.0 on my G4 PowerBook, or my G4 MacMini, and have not yet installed it on my G4 PowerMac.

    Edit: It could be that work on completing any video card driver improvements were stopped and the available manpower (limited) was shifted to other programming tasks that were required to get MorphOS3.0 released on the G4 PowerBook's instead.

    [ Edited by amigadave 11.06.2012 - 19:07 ]
    MorphOS - The best Next Gen Amiga choice.
  • »12.06.12 - 04:03
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    Well, one might assume that "Pilotes 3d plus rapides" mentioned here in green: pilotes3d.jpg would mean that drivers from "2010 MorphOS 3.0 developer version". It turned to be a false assumption, and that's all there is to it.

    P.S. I'm really curious what would be the performance of actual MorphOS 3D drivers on Efika. Should get my hands on one quite soon, I will post the results here ASAP. Here are the Quake 3 results from 2010 dev version (on Radeon 9250, 64 bit):

    640x480 36,5 fps
    800x600 36,2 fps
    1024x768 34,9 fps

    [ Edited by koszer 12.06.2012 - 12:06 ]
  • »12.06.12 - 06:27
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    All there is to it, Koszer, is that Andreas Wolf is trolling and you're feeding him.

    I dusted off my Quake 3 CD and burned a new copy of MorphOS 2.7 and went on to do some testing. I followed Piru's instructions and did setenv TGLSYNC 0 and used /r_primitives 2 for testing on MorphOS 3.0. I left all Quake 3 settings at their defaults to avoid crying over that Quake 3 settings affect frame rates (*GASP*, *SHOCK*, *HORROR*)

    MorphOS 2.7: 72.8 FPS
    MorphOS 3.0: 104.5 FPS
    Speedup: 44%

    I then tried the same test but this time with sound disabled in Quake 3:

    MorphOS 2.7: 84.6 FPS
    MorphOS 3.0: 122.0 FPS
    Speedup: 44%

    Now enter Andreas Wolf crying that 122 FPS is less than 140 FPS, but the fact is that for Quake 3 MorphOS 3.0 is 44% faster than MorphOS 2.7.
  • »12.06.12 - 09:09
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Next up on the list of things to test: RTCW. I don't know of any demo benchmark to do there, so I picked a place on the first map which didn't seem to fast. I saved the game there, and then used the FPS counter at that exact spot to compare MorphOS 2.7 and MorphOS 3.0.

    MorphOS 2.7: 27-28 FPS
    MorphOS 3.0: 51-54 FPS
    Speedup: 90%

    So in RTCW, MorphOS 3.0 is 90% faster than MorphOS 2.7.
  • »12.06.12 - 09:34
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    Thank you for your contribution. Could you please give us some details of your hardware (gfx card, CPU, machine)?

    Edit: I forgot about screen resolution. 640x480, I guess?

    [ Edited by koszer 12.06.2012 - 10:44 ]
  • »12.06.12 - 10:35
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Completely standard 1.5GHz Mac Mini with 512MB RAM.
  • »12.06.12 - 10:42
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > I am not sure I understand [...] the surprise that the video driver
    > in MorphOS3.0 is not as fast as some demostration of an unfinished
    > driver some time in the past.

    I understand very well the surprise that the R200 3D driver is by a wide margin not nearly as fast as was presented in 2010.

    > It was clear to me that the release of MorphOS3.0 would not include
    > the updated video driver(s)

    Then you apparently knew more than us mere mortals.

    > and that 3D support for the PowerBook's Radeon 9700 Mobility video
    > card, as well as 3D support for one or two other currently 2D only
    > supported Radeon video cards, would come in a later release.

    This is another topic altogether, which has nothing to do with the improved R200 3D drivers that were shown in 2010.

    > I guess I must have missed the posting, or announcement from any of
    > the MorphOS Dev. Team members regarding better video card driver
    > performance that was supposed to give us this 140fps performance in
    > Quake III.

    Fab's Alchimie presentation slides list the faster driver as feature for MorphOS 3.0. The 140 FPS is what was shown in 2010, and that very driver was said by the presenters to be included in MorphOS 2.8 (which, as we know, became MorphOS 3.0) back then. Thus I guess it seems natural to assume that the driver from Fab's slides is the one (or an even further improved version of that) from 2010. But seems this assumption was wrong.
  • »12.06.12 - 11:00
    Profile
  • Fab
  • MorphOS Developer
    Fab
    Posts: 1331 from 2003/6/16
    44% and 90% speedups clearly qualify as "faster 3d drivers" to me.

    [ Edited by Fab 12.06.2012 - 10:28 ]
  • »12.06.12 - 11:28
    Profile Visit Website
  • Yokemate of Keyboards
    Yokemate of Keyboards
    magnetic
    Posts: 2129 from 2003/3/1
    From: Los Angeles
    Quote:

    MorphOS 2.7: 27-28 FPS
    MorphOS 3.0: 51-54 FPS
    Speedup: 90%

    So in RTCW, MorphOS 3.0 is 90% faster than MorphOS 2.7.


    wow. Thats impressive. I started to play this game a few months back, but now I think I'll try and finish it! It worked full screen pretty damn fast with 2.7 anyhow..
    Pegasos 2 Rev 2B3 w/ Freescale 7447 "G4" @ 1ghz / 1gb Nanya Ram
    Quad Boot: MorphOS 2.7 | Amiga OS4.1 U4 | Ubuntu PPC GNU/Linux | OS X 10.4
  • »12.06.12 - 11:36
    Profile Visit Website
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    Quote:

    44% and 90% speedups clearly qualify as "faster 3d drivers" to me.


    I guess they do to all of us. Alas, all those who wished their Efikas would wipe the floor with any Pegasos running OS4 will have to wait a little more (darn!) :)
  • »12.06.12 - 12:12
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    koszer,
    Quote:

    Alas, all those who wished their Efikas would wipe the floor with any Pegasos running OS4 will have to wait a little more (darn!) :)


    Are you sure? The new TinyGL in MorphOS 3.0 should support hardware transform and lighting which I would guess helps slow machines such as the Efika much more than it would faster machines such as the Mac Mini.
  • »12.06.12 - 13:02
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > Andreas Wolf is trolling

    I am not.

    > Andreas Wolf crying

    ...is happening nowhere else than in your head alone.

    > that 122 FPS is less than 140 FPS

    You get 122 FPS under the same conditions (hardware specs AND game settings, including resolution) that the 140 FPS were reached under during the 2010 presentation? This would be crucial for a meaningful comparison, wouldn't it?

    > Speedup: 44%
    > [...]
    > the fact is that for Quake 3 MorphOS 3.0 is 44% faster than MorphOS 2.7.

    Another fact is that the speedup shown in 2010 was 200%.
  • »12.06.12 - 13:40
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > 44% [...] speedups clearly qualify as "faster 3d drivers" to me.

    And to everyone else I bet. It's just that a 44% speedup is below the 200% speedup expectations that were stirred by the 2010 presentations. That's btw not whining, crying or complaining, opposed to what some participant of this discussion is trying hard to suggest, but just a factual comparison of hard and cold numbers of present performance vs. numbers that were expected due to things shown in the past on one hand, and curiosity as to what might have been the reasons for withholding the shown driver and as to whether it might be released in the future on the other hand.
  • »12.06.12 - 13:45
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Exactly as predicted, Andreas is back to whining that I only got 122FPS and not 140FPS, being busy calling the 15% difference a wide margin. Furthermore he is now claiming that he was promised to get 253.8FPS, a new wild claim. I guess that every time Andreas is proven wrong, he has to make up some new outrageous claim to continue to have something to cry about.
  • »12.06.12 - 14:30
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    Thanks for the hint, so now it's time to see for myself if that's true (as soon as I get my Efika back).
  • »12.06.12 - 14:31
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    @Wolf_Andreas:

    Oh please, you were first to complain that the contitions weren't comparable. You compare your 122 FPS (at 640x480) with 140 FPS (archieved at 800x600). Now that ain't comparable.

    [ Edited by koszer 12.06.2012 - 14:38 ]
  • »12.06.12 - 14:36
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Quote:

    You compare your 122 FPS (at 640x480) with 140 FPS (archieved at 800x600).


    Why are you so sure that the 140 FPS was at 800x600?
  • »12.06.12 - 15:01
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > Andreas is back to whining

    And you're still mistakenly taking my corrections of your nonsense and my questions as whining.

    > that I only got 122FPS and not 140FPS

    False. I was asking you if both those results were achieved under comparable conditions like hardware specs and game settings. You chose to not answer that question, which speaks volumes.

    > being busy calling the 15% difference a wide margin.

    Wrong again. With "wide margin" a was referring to the difference between 2010's 140 FPS and koszer's reported 78.5 FPS under comparable conditions, not your 122 FPS.

    > he is now claiming that he was promised

    In your head. I've not even used the word "promise", liar.

    > to get 253.8FPS, a new wild claim.

    By you that is, Mr. liar.

    > I guess that every time Andreas is proven wrong

    ...which happened not even once in this thread.

    > he has to make up some new outrageous claim to continue to have
    > something to cry about.

    The only one here making up stupid claims is you, as you keep proving with each and every posting you utter.
  • »12.06.12 - 15:11
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Good Lord, Andreas, you sure are quick at making up bullshit. Not only have you decided that your guess as to what hardware is faster for running Quake 3 is the one true truth, you have also decided that your guess as to what Quake 3 settings were used is absolute gospel, and you have decided that my results are invalid because I specified exactly what I did and on what hardware, thus not being accurate enough for you, while your own wild imagination is good enough to be considered absolute truth. Because one person is reporting being unable to get higher FPS, quite likely because of random modifications done to said person's system or setup, you discard all reports of things working exactly as expected. Can't you for once in your life realise that all your spamming was for nothing, and all you're doing by continuing your crying parade is spam everyone with useless messages?

    Seriously, how about you start up that blog so you can make all the sensationalist messages such as "OMG, I swear, some guy reports that Quake 3 on MorphOS 3 only achieved 4 FPS(*), which is less than the 666FPS promised 2 years ago. (*) system was running in unregistered slowdown mode."?
  • »12.06.12 - 15:20
    Profile
  • Paladin of the Pegasos
    Paladin of the Pegasos
    koszer
    Posts: 1246 from 2004/2/8
    From: Poland
    Quote:

    Why are you so sure that the 140 FPS was at 800x600?


    Not sure, just quoting guruman's post from 15.11.2010, that says:

    Then we were shown the updated 3D drivers, reaching 140fps in Quake III running at 800x600 on an humble Mac mini. That's roughly 50% better than last year, that in turn was already 100% faster than what we have in the public version of MorphOS...
  • »12.06.12 - 15:21
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    Quote:

    MorphOS 2.7: 84.6 FPS
    MorphOS 3.0: 122.0 FPS
    Speedup: 44%


    Quote:

    It's just that a 44% speedup is below the 200% speedup expectations that were stirred by the 2010 presentations.


    Andreas, 84.6 FPS + 200% = 253.8 FPS. Don't claim you didn't say this. That just makes you look retarded, rather than trolling.

    [ Edited by Wolf_Andreas 12.06.2012 - 14:24 ]
  • »12.06.12 - 15:22
    Profile
  • Cim
  • Just looking around
    Cim
    Posts: 9 from 2012/6/9
    I got 148.1 fps in 800*600 on Mac Mini :P
    How? I set all graphics options to minimum, sound off, all game options to off.
  • »12.06.12 - 16:43
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > Andreas, 84.6 FPS + 200% = 253.8 FPS. Don't claim you didn't say this.

    But I do, see: I didn't say this. The only one of those figures I mentioned here is the "200%" one, which is what was reported in 2010:

    "That's roughly 50% better than last year, that in turn was already 100% faster than what we have in the public version of MorphOS..."

    Do your math and you'll see that this would amount to a 200% improvement, as (1+1)*1.5-1 = 2.
    Now we know that the frame rate after this 200% improvement was reported as 140 FPS. Calculating the base value, we can see that the frame rate with the older released driver must have been (around, as 140 is probably a rounded figure) 140/3 = 46.7 FPS.
    Now, as to why you get as much as alleged 84.6 FPS under alleged similar conditions (hardware specs AND game settings) others obviously got only 46.7 FPS (i.e. little more than half of your result) under, I don't know. Do you have an explanation?
    Fact remains, and can easily be seen, that both FPS figures "84.6" and "253.8" came from you alone and your claim those were my figures is a lie.
  • »12.06.12 - 16:47
    Profile
  • Yokemate of Keyboards
    Yokemate of Keyboards
    Andreas_Wolf
    Posts: 12134 from 2003/5/22
    From: Germany
    > Andreas, you sure are quick at making up bullshit.

    As for example?

    > you decided that your guess as to what hardware is faster for running
    > Quake 3 is the one true truth

    Huh? Care to quote me on that?

    > you have also decided that your guess as to what Quake 3 settings
    > were used is absolute gospel

    Huh? Care to quote me on this alleged guess of mine? The only one setting I ever mentioned here is the screen resolution of 800x600, which is what was reported back then by others, not by me. Do you think those reports are wrong?

    > you have decided that my results are invalid

    Huh? Care to quote me on that? I merely asked you whether your hardware and game settings were comparable to what was used in 2010. Are they? Yes or no? Or will you dodge this question once more?

    > because I specified exactly what I did and on what hardware

    Where and when did you specify your screen resolution?

    > thus not being accurate enough for you

    The deviations we're talking about here are far too great to be the result of missing accuracy.

    > while your own wild imagination is good enough to be considered
    > absolute truth.

    Care to quote me on my "own wild imagination"?

    > Because one person is reporting being unable to get higher FPS

    It was 3 (verbal three) persons, Mr. liar: redrumloa, Cim and koszer.

    > quite likely because of random modifications done to said person's
    > system or setup

    So you say redrumloa, Cim and koszer made an arrangement of some obscure kind?

    > you discard all reports of things working exactly as expected.

    Care to point out all those reports that are reporting a Quake 3 frame rate of about 140 FPS at 800x600 resolution on a 1.5 GHz Mac mini and with game settings that resemble those that can be assumed from the 2010 videos? I surely must have missed them.

    > Can't you for once in your life realise that all your spamming was
    > for nothing

    As this thread has proof of, my "spamming" as you call it was not for nothing.

    > all you're doing by continuing your crying parade

    There is it again, your fantasy about me crying. You can't just let it go, can you?

    > is spam everyone with useless messages?

    That's what you have been doing, Mr. liar, right from the beginning.

    > you can make all the sensationalist messages such as "OMG, I swear, some guy
    > reports that Quake 3 on MorphOS 3 only achieved 4 FPS(*), which is less than the
    > 666FPS promised 2 years ago. (*) system was running in unregistered slowdown mode."?

    Make that "2 guys report that Quake 3 on MorphOS 3 only achieves 50 to 80 FPS, which is less than the 140 FPS reported 2 years ago", and it makes even sense.
  • »12.06.12 - 17:24
    Profile
  • Caterpillar
    Caterpillar
    Wolf_Andreas
    Posts: 23 from 2012/6/11
    So how is that blog coming along?
  • »12.06.12 - 17:41
    Profile