man... i wish apple would host this download, my connection is normally pretty quick,,,, but wow. this d/l is going to take all day.
Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Unreal Tournament 2003 Demo Released
- Thread starter MacRumors
- Start date
- Sort by reaction score
well i hear ut2003 uses 1 chip for game 1 for sound. still a great loss when using 2. this is why a single 1.4 970 will blow the doors off a 1.42 dual g4. lots of loss potential when trying to get programs to use both cpu's at 100% aint going to happen but now you have a single chip executing more instructions at the same clock. Upgrade my powermac or bite the bullet and get a new 970. we are about 1 month away folks. The 970 will be a new level of performance that will only get better.
Well according to the developer utilizing both chips did see an increase in frame rates. Though as I have stated on a high end machine a dual/1.42 each cpu was only seeing 60% load and that includes system operations and everything else. I this shows that it isn't as simple as a loss to the processing to distribute work load but it is also a lack of efficiency on the programs side otherwise I would think it could achieve at least 80% or more load on the cpu that is being used to process the game engine.
UT2003 a good example of hardware acceleration
I disagree that the Unreal Tournament 2003 application is badly written because the processors are not being used 100% each. It seems like all the hard work is done by the graphics card and the CPUs are utilised for things like sound and physics; isn't this the way console games utilise their limited host hardware? I think in the long-term this way of doing things is better for us Powermac owners because a graphics card upgrade would be cheaper and increase our framerates more than a CPU upgrade would.
I disagree that the Unreal Tournament 2003 application is badly written because the processors are not being used 100% each. It seems like all the hard work is done by the graphics card and the CPUs are utilised for things like sound and physics; isn't this the way console games utilise their limited host hardware? I think in the long-term this way of doing things is better for us Powermac owners because a graphics card upgrade would be cheaper and increase our framerates more than a CPU upgrade would.
Re: UT2003 a good example of hardware acceleration
No, I agree this is the way it should be. Though I don't think this is 100% of the case. If this were the case going from a G4MX to a ATI9700 should yield an almost 4x increase in frame rates. It doesn't. It's not even a 25% increase in frame rates.
No, I agree this is the way it should be. Though I don't think this is 100% of the case. If this were the case going from a G4MX to a ATI9700 should yield an almost 4x increase in frame rates. It doesn't. It's not even a 25% increase in frame rates.
you have to feed the graphics card, best graphic card in the world isnt going to help you if you are running a 733 like me UT2003,Halo,and Doom3 are going to be pushing the limits on newer systems. i wonder if anyone can run nascar 2002 with full track and everything on? it will be nice to see a 970 run this stuff everything on & more.
CPU speeds
I have UT2K3 on my PC at home and at work (what's it doing there???
) at work I have a 1.7 GHz P4 with a 64Mb Radeon 9000 and I pull about 30fps. At home I have a 1.4 GHz Athlon with a 128Mb Geforce 3 Ti 200 and I pull about 40 fps. The CPU definately does help, but so does the GPU.
* On a side note, I just bought my first song off the Apple Music Store! WOOHOO! Go Apple Go!!!
I have UT2K3 on my PC at home and at work (what's it doing there???
) at work I have a 1.7 GHz P4 with a 64Mb Radeon 9000 and I pull about 30fps. At home I have a 1.4 GHz Athlon with a 128Mb Geforce 3 Ti 200 and I pull about 40 fps. The CPU definately does help, but so does the GPU.* On a side note, I just bought my first song off the Apple Music Store! WOOHOO! Go Apple Go!!!
believe it or not there was. it's no faster than the 64Mb Geforce3 but I just wanted to see the 128Mbs of ram on the video cards boot screen It was the same price as the 64Mb so why not?
It was the same price as the 64Mb so why not?
Originally posted by ExoticFish
believe it or not there was. it's no faster than the 64Mb Geforce3 but I just wanted to see the 128Mbs of ram on the video cards boot screen
Ah, but it should be faster with these newer games that have larger textures as it will prevent having to load additional textures as the game progresses.
I agree that with the PC version there is a substatial noticeable gain by going from one video card to the next but I haven't seen proof of that with the Mac Demo version yet. In fact I have been able to achieve faster frame rates with a couple tweaks at nearly double the screen resolution with a G4MX 32MB then an ATI9700 or Ti4600 in the same computer as mine (Dual/GHz/DDR PowerMac) with the base settings in the bot match benchmarks.
It's most definately obvious that the CPU is what's holding this game back, not the video cards, I agree, just saying that every little bit helps. and I don't know what I was smoking, the video card in my machine at work is a Radeon 7500 not a 9000. A 9000 would be much better.
How can you say that when every cpu load meter that I have seen shows no more then 60-70% of cpu useage while running the game. Also there is very little difference in frame rates between a 700MHz G4 up to a Dual 1.42GHz G4 with any graphics board in them. It has to be the software.
Well 70% of a 733MHz CPU is less than 70% of a Dual 1GHz system. My friends dual 1 GHz machine gets about 10 more fps than my single 1GHz TiBook. I'm not saying that I don't agree with you. I'm saying that factors on every front contribute to the sluggishness of the game. I.E.
1) The game is not optimized enough
2) The G4 is long due to be replaced
etc.... etc....
But you saying that the frame rates don't vary that much between 700MHz and a Dual 1.42 GHz machine is a little unrealistic.
1) The game is not optimized enough
2) The G4 is long due to be replaced
etc.... etc....
But you saying that the frame rates don't vary that much between 700MHz and a Dual 1.42 GHz machine is a little unrealistic.
They don't vary much. If the cpu was the only thing holding the game back as you had suggested then you should see a nearly linear increase of frame rates with CPU speed. That said if a 700MHz G4 is achieving 30FPS on average a Dual 1.42 should achieve around 55FPS even if only one of the CPUs is doing the work. That simply isn't the case. I haven't seen any Mac of any sort with any video card achieving an average of 55FPS on a botmatch benchmark. The reason your buddy with the dual is achieving better frame rates is that all the audio rendering is off loaded to the second cpu. Also the main game can occupy the cpu without the system and finder and any other app running on it. You can even out the differintial between your machine and the dual by simply cutting the sounds channels down from the standard 32 to around 12. This should help greatly on any single processor machine.
It boils down to framerates and to get em where we would love them you have to have both cpu power and gpu power. one without the other isnt going to help much. And we all know the g4 has been lacking, and apps are not written to use it at its max. So here we are. and yes frame rates between a 733 and dual 1.42 are going to be very different. Anyone who says they are not has just burnt one and is euphoric.
Just read the results your self either in this thread or over at Xlr8yourmac.com. The botmach benchmarks between a high end system and a mid-weight system such as a 700-800MHz G4 have less then a 15% margin between them.
come on bandit i was just there what graph are you talking about? i saw where my 733 was getting 30 fps in wolfenstein and a single 1.2 upgrade was getting 50. 20 fps can be the difference of constant smooth action or stuttering slow down also those same 20 fps may allow you crank up to the next resolution depending on your gpu and play at higher resolution or more things on with a exceptable slow down. When compared to the pc world our gaming frame rates suck. low clock g4s
We're not talking about Wolfenstein are we? We're talking about the frame rates achieved in UT2003.
My point is that it's not the low clock G4s getting us low frame rates. Go look at any QuakeIIIArena frame rate comparrisons and you will see that a Dual processor G4 will compare favorably against all but the very very newest PC systems. The problem is not the computer it's the software.
Well Quake 3 was programed with the Mac in mind. It's probably the most optimized game out for Mac! But UT2K3 is a port. Plus, I can pull over 100 fps in Quake 3 on my PC while only 40 in UT2K3, and that's not even all the time! Not sure what fps I get on my Ti w/ Quake 3 but I know that it's kick ass too. UT2K3 is slow because it needs power that isn't going to be available till the 970's show up. UT2K3 is slow on any machine except the fastest available on BOTH platforms. Same thing happened when the first Unreal showed up. Not a lot of people played it untill faster computers got cheaper.
Quake3 was optimized for mac and a dual will slaughter a single cpu, It doesnt matter wolfenstein,ut2003,doom3 on and on how about medal of honors beach landing? cpu speed and programing, but we as consumers can only vote our wallet. So as a Mac gamer we want higher cpu speeds since we dont program these games. I wont argue the g4's lack of speed since this is a known fact and is why powermac sales have hit rock bottom.
Macbandit is completely correct - the software is a dog at this stage.
There's no point arguing about G4s etc - a game of this complexity (which isn't that much higher than the previous gen) should be able to pull a much higher framerate on today's CPUs and GPUs. I'm pretty convinced it's due to a lazy port to OpenGL.
As a subsidiary point, I know the framerates of the original UT were a constant source of complaint, but Westlake made sure that it ran well enough even on hardware that was quite old - it didn't even require harware graphics acceleration for God's sake! The beauty of that game has been completely lost in this new version - PC-using friends who loved the original UT agree.
There's no point arguing about G4s etc - a game of this complexity (which isn't that much higher than the previous gen) should be able to pull a much higher framerate on today's CPUs and GPUs. I'm pretty convinced it's due to a lazy port to OpenGL.
As a subsidiary point, I know the framerates of the original UT were a constant source of complaint, but Westlake made sure that it ran well enough even on hardware that was quite old - it didn't even require harware graphics acceleration for God's sake! The beauty of that game has been completely lost in this new version - PC-using friends who loved the original UT agree.
I think the original was the better game, sure we have new bells and whistles but the original was the ticket that rocked my gaming world for the longest of any game.
Definitely. They have a lot of work cut out for them to even duplicate the playability of the first one let alone make it run on a variety of machines and still have decent graphics. That's one of my biggest gripes about UT2003. You turn the graphics down to nothing and you get something that looks worse then Doom or Marathon and it still doesn't perform better. What's with that?