MacWorld posted some basic benchmarks of Doom 3 on the Mac.
Doom 3 for MacOSX benchmarks at Macworld.com
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
Geez … 47.1 at 640×480 is the *best* score?
Cripes. Guess that’ll keep gamers on WinTel.
He said, “Games don’t work that way, though: there’s no substitute for CPU cycles when it comes to games like this.”
I’m not sure what he meant there – Athlon’s are able to outperform higher clocked Pentiums no problem.
please keep in mind that they used “High Quality” setting…also, from the article: “Aspyr has also benchmarked a Power Mac G5 1.8GHz system equipped with an ATI Radeon 9800 Pro Mac Special Edition card — considered a “mid-range” system for this game — operating at about 29.6 frames per second, compared to 32 frames per second for a similarly equipped Dell.”
so the difference isn’t that big…with future optimizations (Tiger…) we may see this difference shrink even more. (or even see the Macs pull out ahead, as we saw it with Quake 3 when Altivec optimizations were added)
“…we are seeing a very playable frame rate of 29.5 FPS running at 1600 x 1200 pixels.”
Wrong! very wrong! I get around 50fps in that timedemo and the game is VERY slow on the BOSS levels with lots of monsters on the screen. The timedemo needs to run around 60 fps in order to get a smooth play through the whole game. 29.5 is not gonna cut it, but it will work for the first few levels.
I don’t think there is anything inherently wrong with the Mac platform in terms of gaming performance. I’m developing a cross platform open source driving simulator ( http://www.motorsport-sim.org )for Win/Mac/Linux and while we have it compiling in Linux/Windows we can’t find anybody to take the few hours to make a Mac binary. I’m hoping that a game based on OGRE, a cross platform graphics engine, becomes the standard for testing performance differences between operating systems.
An article containing SOME substance being published on MacWorld. I can’t believe it. Usually they are merely a parrot for Apple.
I almost laughed at the benchmarks at first, returning numbers 70% what my Barton 2500 with a mere Radeon 9500 pro returns, however I know the Mac platform always falls short in the realtime graphics arena…
In part from it’s piss poor tasking model. As the article said OSX as a rule doesn’t like to hand off full time control of resources to a program, their priority model costing you precious response time.
I oft find it amazing how a OS known for not multitasking well also does not single task well either… I think it’s probably why when I’m stuck on a Mac I feel like the programs are running slower than they actually are… Execution times return faster, but you’d never know it from the programs responses to inputs or the screen redraws.
That’s why I have a PS2. Gaming is better on consoles anyways.
couldn’t have said it better
i wish that were true, but it just isn’t. my wintendo 2ghz opteron with a 6800 ultra gets about 60fps at 1920×1200 (the cinema display res). i love my powermac, but it wouldn’t get anywhere close to that with this half assed port
. the reason quake3 ran so damn well is one of the core developers took a personal interest in optimizing it for the mac. sadly that isn’t the case for doom3.
i think you don’t have a single clue what the hell you’re talking about. go look up the benchmarks sometime for quake3. then try and explain why macs with identical video cards to pcs manage to get higher framerates. macs are just fine for graphics (if not better since opengl part of the core system so it’s heavily optimized). doom3 was poorly ported, it’s as simple as that.
I used to think the same until I got broadband and a graphics card.
You’re all aware that this game has a cap of 60 fps, right ?
please keep in mind that they used “High Quality” setting…
Most, if not all, benchmarks on the Windows platform used high quality also.
also, from the article: “Aspyr has also benchmarked a Power Mac G5 1.8GHz system equipped with an ATI Radeon 9800 Pro Mac Special Edition card — considered a “mid-range” system for this game — operating at about 29.6 frames per second, compared to 32 frames per second for a similarly equipped Dell.”
That’s not the case. The graphics card is basicly the same as the PC version except for the BIOS, ATI’s OpenGL drivers suck (on both platforms :p) we already know that.
Aspyr tells us they worked diligently with Apple to make sure “no stone was left unturned” when it came to optimizing Doom 3 for the PowerPC’s Velocity Engine registers, so you’ll be happy to know that Doom 3 is well-tuned for your G4 and G5.
I somehow doubt they’d make that statement without it beeing true as this game is surely something Apple would most likely help to get it running well, because of it could (can/will) attract more users to buy a Mac.
so the difference isn’t that big…with future optimizations (Tiger…) we may see this difference shrink even more. (or even see the Macs pull out ahead, as we saw it with Quake 3 when Altivec optimizations were added)
Sadly, the differene is big. Remember they used a dual G5 2.5 GHz. That’s a beast compared to my current Athlon 2600+ (2.13 GHz) with a Radeon 9800pro (128 bit memory I believe). Also they used the X800 XT which, of course is a beast to my card. They also used 3 times the RAM my PC currently possesses. And I get about the same FPS this guy is getting (best I remember I got 47 FPS in timedemo – nothing overclocked, running Windows XP).
Doom is also very heavy on CPU usage – all the lightning is done on the CPU for example, as well as sound, physics, AI,… So theoreticly, that Mac ‘should’ have scored _a lot_ higher then my setup.
i think you don’t have a single clue what the hell you’re talking about. go look up the benchmarks sometime for quake3. then try and explain why macs with identical video cards to pcs manage to get higher framerates.
Well no replace ‘Quake 3’ with ‘Doom 3’ and you see the results. From what I can tell (and I’m no expert) this “problem” (if it can be called like that) can be nailed down to 2 things: (a) CPU or (b) the OS. Either the CPU architecture (obviously not the speed) is not well fitted for the Doom engine (remember that lightning is done on the CPU) or the OS has (possibly) problems with singletasking. If you look at this article -> http://www.anandtech.com/cpuchipsets/showdoc.aspx?i=2149 you can see that Doom 3 is almost entirely CPU limited.
Though I do not claim it is because of the OS or the CPU all evidence points to it and I, myself am very confident it is something to do with the OS.
btw, about Q3 on the Mac. Did you try to test it on 10.3.8 – that’s what they used and that’s what D3 needs on the Mac. Though this wouldn’t prove much because Q3 is not as nearly as CPU dependent as Doom 3.
btw, hoover, not in timedemo.
I wonder how much of this can be attributed to compiler optimization. I’ve always been of the understanding that the GCC compilers for PPC have lagged behind the ones for x86 and Microsoft’s with regards to application performance and optimization.
Yes, I ad forgotten about that. Also, the doom3 engine does not take advantage of dual CPU setups like the quake3 engine did.
The bottom line from the article:
If you were hoping that your dual-processor G5 was going to be a Pentium or Athlon-killer when it comes to Doom 3 framerates, you’re going to be disappointed.
The current state of game programming IS multiple processors. Shame on you Carmack.
Xenon’s last rumored to have 3 CPUs on a die. PS3 is going to have quite a few as well.
Their game engine better start supporting multiple processors or it’s going to get left behind.
This is the saddest thing ever, i get higher frame rate in Linux with a 9800 non pro at up to 1024×768. This just shows how cheap apple is, If you try and buy one of there mac’s for $2000 you will get a good nvidia 5200 isn’t that nice. This is for $2000! Also they kept comparing similarly equipped pc’s by dell, why not try similar prices by dell, because i know those macs didn’t cost $800-900 and the dells probably had 1.8ghz celerons knowing dell.
you apparently did not read about the test rig. it was running a X800 with 256 MBs of RAM.
when they turned off the 3d rendering it got 106 FPS which points to the Graphics engine as the slow down. and this again can be pointed to the fact that the Mac is a better Multitasker than the PC and the PC is a better old school, low tech, single tasking system.
when they turned off the 3d rendering it got 106 FPS which points to the Graphics engine as the slow down.
I think you got there something. You get higher FPS if you disable the rendering of 3D objects. Better mail this to id so they put it in their next patch
His point, you seemed to miss, is that the bottleneck is OpenGL and possibly the video drivers, NOT the CPU as everyone seems to be having the typical G5/AMD/Intel ‘mines faster than yours’, ego trip again. :/
Kevin you wrote:
His point, you seemed to miss, is that the bottleneck is OpenGL and possibly the video drivers, NOT the CPU as everyone seems to be having the typical G5/AMD/Intel ‘mines faster than yours’, ego trip again. :/
You are correct on the video device drivers. I/O Team doesn’t write the driver for the X800. They write the video drivers for only the graphics cards offered at the http://store.apple.com.
This means that you must rely on ATI to have a team who are well-versed in I/O Kit’s C++ Architecture.
I’d expect this to change when Tiger is released and the updated Kits are presented with CoreImage.
His point, you seemed to miss, is that the bottleneck is OpenGL and possibly the video drivers, NOT the CPU as everyone seems to be having the typical G5/AMD/Intel ‘mines faster than yours’, ego trip again. :/
I don’t think i missed his point. Below is the exact snippet from the article:
Running the same demo at 640 x 480 with 3D rendering turned off — which can be done by setting the r_skiprender value to 1 in the Doom 3 console, then running the timedemo — returns an FPS average of better than 106 frames per second. Setting that command gets all the 3D stuff out of the way, telling us how fast the core game engine itself is running. And that number compares pretty well from Macs to PCs. That implies that the performance difference we’re seeing has something to do with the way the 3D graphics are rendered, as opposed to any specific shortcoming in the game engine or overall slowdown that would account for a difference.
r_skiprender skips rendering of 3D but keeps 2D, basicly you are left with the GUI and some other irrelevant stuff. I said in my previous comment (check one of Carmack’s .plans to verify) that all stencil shadows are beeing done on the CPU. What they did was disabled all 3D rendering which means no polygons are beeing draws, so now shadows are beeing calculated thus freeing up the CPU.
good luck playing Doom3 on your PS2!
hoover: sure, but that doesn’t mean having a system capable of producing more than 60fps on average is useful. An average doesn’t tell the whole story; a card that produces 60fps average may only manage 20fps when the screen gets busy, one that can do 120-150fps average will keep at least 60fps all the time.
A lot of it comes from the card. Well, I did notice that they were using the latest ATI card. Well, there is still one disclaimer:
“All tests were performed at “High Quality” video settings with all Advanced Options turned on (High Quality Special Effects, Enable Shadows, Enable Specular, Enable Bump Maps) except for Vertical Sync. Anisotropic filtering was set to 8x. Frames per second averages were achieved running the timedemo demo1 command; the demo was run twice, and the higher average was used.”
I would like to have seen this test system in low quality mode.
From the article:
“Running the same demo at 640 x 480 with 3D rendering turned off — which can be done by setting the r_skiprender value to 1 in the Doom 3 console, then running the timedemo — returns an FPS average of better than 106 frames per second.”
Seems like the bashing is coming from people who don’t read the article.
FYI: I don’t own, use, run a Mac. But I believe in giving credit where credit is due. And I don’t play games (any more).
Enjoy
Sorry Hoover,
“Running the same demo at 640 x 480 with 3D rendering turned off — which can be done by setting the r_skiprender value to 1 in the Doom 3 console, then running the timedemo — returns an FPS average of better than 106 frames per second.”
This is from the article. How did they hit 106 if there is a cap @ 60 FPS.
In the end guys the “Average Joe” on the street is not gonna tell the diference anyways. It’s just the usual chest beating between Wintel and Mac users…
I wasn’t talking to much about the high end system, really what i was talking about is how they compared a dell at what they said was comparable system to the imac g5, the problem with what they stated was that they never said what type of system or how much cheaper it is, it could have been a 2-3 years old and just slapped a 9700pro in it. I am not saying the g5 is a bad cpu. What i really hate apple for is giving people $2000-2500 systems with 512 mb of ram and a nvidia 5200 graphic card, This is what i was talking about when i said cheep. Also like many have said it probably was the video card drivers, but apple should never be selling ati cards in their systems since ati has a bad record compared to nvidia in opengl.
With all the features. What more can you ask for?
For those of you who are uninformed …
The Doom 3 60 FPS cap is only during gameplay. A timedemo runs the game at the highest speed the hardware can push.
http://www.neoseeker.com/Articles/Hardware/Previews/atix850p/9.html
Looking at the X800 scores there, it’s apparent the Mac gets pwned — and this is running on a P4 3.4EE with 1 GB of RAM.
Amazing… a computer running at 3.4 Ghz outperforms a computer running at 2.5 Ghz playing a video game.
Of course, for a CPU (P4 3.4EE) that comes with 2MB of cache built-in and costs around $1000 just by itself, I would hope it could put up some good numbers.
But what about the 2.4 ghz athlon 64 that crushes both the 3.4 P4 and the 2.5 g5? Your logic says that higher mhz always wins in games but you’re obviously wrong, its the g5 architecure that is mediocre at games, it’s no A64.
“Amazing… a computer running at 3.4 Ghz outperforms a computer running at 2.5 Ghz playing a video game.”
You could have tried to compare whith an Athlon 64 FX55 which runs at 2,6Ghz instead.
(The FX55 equals an overclocked PIV (660) running at 5Ghz when it comes to video games : http://www.hardwareanalysis.com/content/article/1787/).
So I don’t think the comparaison is unfair.
is that it doesn’t matter if your gfx card can do all the fx the game engine allows or not because the game is just too dark for you to see them :-P, plus the AI cheats.
Seriously, halflife2 or even older games like no one lives forever 2 are more fun to play.
I’m still going to try it on my 1GHz Powerbook with a wonderful FX5200 w/ 32MB!
He is talking about that you cant beat Ghz when it comes to gaming.
BUT Why is it then that the Athlon FX is undoubtly the fastest gaming machine out there?? It has 2,6 Ghz whereas the P4 has 3,4 or something ( i may be wrong on the actual numbers, but you get the idea )
Cheers
Can anyone see what their fps is in windows when you turn off the graphics like they did? I hate it when an article says two scores are comparable, but doesn’t actually tell you what the 2nd score is… I really think the graphics drivers are the bigget issue here. That’s why the scores in Linux are lower than in Windows even on the same hardware.
And why do people keep talking about how many MHz the P4 is running at? That P4 would get owned by a 2GHz A64, let alone one running at 2.5GHz like the Mac.
You could have tried to compare whith an Athlon 64 FX55 which runs at 2,6Ghz instead.
(The FX55 equals an overclocked PIV (660) running at 5Ghz when it comes to video games : http://www.hardwareanalysis.com/content/article/1787/).
So I don’t think the comparaison is unfair.
Wow. These benchmarks are also all showing the performance of PCs with PCIe graphics cards. Obviously, because of the greater bus throughput, the latest games are going to show better performance on a PC with PCIe than a Mac which currently only has AGP (even if everything else was equal). So combine this with a ported codebase which I’m certain we’ll see further optimization to, along with better optimized display drivers (which come with every point release of OS X) and compilers, I don’t think the Mac will be sucking hind-tit for too long.
Of course I also don’t know how many people bought their Macs so that they could brag about the performance of Doom 3 while they waste enough electricity to light a stadium.
I don’t think the Mac is bad at games. But you people have to stop talking about the great performance you get in Quake 3. It’s so old that right now it’s nothing more than a bandwidth test. Macs are good at this game because they have a fast FSB.
And right now there is absolutely no difference in speed between AGP cards and PCIe cards. In fact, there’s not much difference even when you go down to AGP 4x.
well, no surprises for me.
looking the quacke 2 code I see tat ID make a very good optimization for x86 proc, but unlucky is traduced in very poor optimization for ppc. so performances are at best in pair with x86 per clock. plus, the apple opengl is not very fast for games, so another performances problem. amd don’ t forget audio acelleration, a good audio card is needed for gain fps.
I get noticeably worse performance on Doom 3 on Linux than on w2k. One of the id guys said it is because gcc has worse SSE/MMX optimization than MSVC. As I understand it x86 is the main development platform for GCC, so G5 optimizations are probably not even as good.
Supposedly GCC 4 will be ultra super hyper fast. We’ll see.
I am a Mac/PC user and I can say that the Mac OS X has big graphics problems. Apple is always promising new graphics and interface response performance boost but it neves reaches a mid Wintel machine. My Athlon+ 2800/1GB is faster than my Dual G5 2.5. G5 is faster in interface response but the graphics processes are still slower than PCs. I think OSX loses performance in some place that Apple cannot change and the company is waiting for bigger and faster processors to hide this issue.
It looks like this will struggle to work a Mac Mini (low quality with dynamic shadows turned off) and, which is more shocking, on a new iMac.
The new iMacs have a FX5200, which people like myself complained about at the time. I used have that card on my PC. The timedemos posted roughly 35fps with shadows turned on, however during gameplay it would drop into the teens during complex scenes with lots of entities and / or characters. As a result, to get a decent frame-rate, I played on medium quality at 640×480 with dynamic shadows turned off[1].
The fact of the matter is Apple just don’t make computers for gaming.
[1] A lot of people seem to get confused about the lighting in Doom 3. Dynamic shadows are only used for entities and characters. The shadows cast by the architecture (which is unchanging) are pre-calculated, just like previous titles. If you turn off dynamic shadows the game still looks good, it just loses some atmosphere (the designers used shadows for some eerie moments early on in the game)
” The shadows cast by the architecture (which is unchanging) are pre-calculated, just like previous titles.”
Are you sure about that?
I’m pretty sure you get it wrong. Dynamic lighting (for everything) is supposed to be the big difference between Doom 3 and its competitors!
Try to play with the editor : moving brushes (or objects or whatever you want, even harchtecture) changes the casted shadows in real time (in the real-time preview window).
That’s why doom 3 the engine is so demanding towards hardware.
Wow! So much non-sense on so tiny space.
Apparently gamers don’t know much about technology. As long as a system has the right buzz-word compliant components, it must be fast, right? And if a system does not performe well with one app, it must be a hardware or os problem (even if the same hardware / os perfomerm well in other benchmarks).
And how really, really, really important is it to have the fastest of the fastest even if the same thing is a snail within less than three months…?
Get a life..
—
The reason computer chips are so small is computers don’t eat much.
I recommend Far Cry for all of the action fanatics, has way better video engine, and on my pc worked much better than doom 3 did. First i played Far Cry and everything went smooth and fast on my radeon 9500 pro, then i ran doom3 and went “what’s that slow shit”? and this is on a pc with 786MB of ram. come on. it gets another 512 mb not to read from disk for this game……talk about optimization.
and there’s not so much of the new technology in this game either…a lot of things are faked using textures (like metal crates for instance)and the faces are where? most of the game is corridors walking…i think its way worse than crytek’s engine achievement. not to mention the new dimension of gameplay they achieved.
i think doom3 is badly optimized even as a pc game, and relies heavy on the compiler optimizations. as for the mac, no wonder this turned out that way.
another engine for mass duplication on the mac, please.
talk about “when it’s done” commitment
DOOM 3 is a JOKE!
Doom3 works fine on my 5200 pci 256MB, however Far Cry chokes.
Doom 3 has a very powerful engine….and VERY demanding…
we wont be able to see its full capabilities until at 1 more year….
[1] A lot of people seem to get confused about the lighting in Doom 3. Dynamic shadows are only used for entities and characters. The shadows cast by the architecture (which is unchanging) are pre-calculated, just like previous titles. If you turn off dynamic shadows the game still looks good, it just loses some atmosphere (the designers used shadows for some eerie moments early on in the game)
Sorry, but you are wrong here. Doom uses stencil shadows for all 3D objects. Meaning full dynamic lightning/shadowing all the time. It uses caching for unchaning lights because they don’t need to be recalculated every frame. Quake 3 used stencil shadows for entities you can try that with r_shadows 2 (or 3, can’t remember).
I get noticeably worse performance on Doom 3 on Linux than on w2k. One of the id guys said it is because gcc has worse SSE/MMX optimization than MSVC.
Is has been said (by the same person who posted that .plan) that if he would have optimized it for gcc it would gain like 8% of speed.
Just curious why they didn’t compile doom3 with Intel’s compile or Portland Groups compiler. I figure because they are developing a commercial game they should be able to afford one liscence fee.
> Just curious why they didn’t compile doom3
> with Intel’s compile or Portland Groups compiler.
Wild speculation by me:
1) Gcc is very standards compliant, which is good for SW development perspective.
2) Porting to Mac (and other systems in general) becomes easier due to widespread use of gcc.
3) id’s Linux support is more of a “hey this is cool” -kinda thing, so using free tools fits in with that philosophy.
It’s slow, really slow… don’t know if the fault is Id-Software or Apple (hardware or OS) but in my opinion there are no exuses for the frame rates obtained.
hello people!!!! read the article. the FPS performance has EVERYTHING to do with the GFX performance and the better multitasking abilities of OS X and nothing to do with the CPU.
they turned off the 3D rendering, they got 106 FPS. the G5 is the only thing in the equation now. that shows the G5 is a good CPU. the 3D graphics performance is not dictated by the CPU at all, that is the point of the GPU for god sakes!!!
I was truly disappointed by these results. The fact that the test machine was the fastest mac that money can buy with loads of RAM and a top-end graphics card makes these mediocre frame rates even harder to stomach.
You get the usual zealot response, blaming every non-Apple link in the chain: ATI drivers, compiler optimisation, etc, etc. At the end of the day, does it really matter whose fault it is? The bottom line is that the Doom 3 experience on the fastest mac hardware falls well short of the mark. That’s a real shame.
didnt read all the comments (the small amount i did read really blow my mind, indeed, gamers know nothing about technology.), but you need to do some config file tweaking or performance is garbage, at least on pc. doom was a real crappy game anyways, you really arnt missing much if you cant play it.
//That’s why I have a PS2. Gaming is better on consoles anyways.//
Have fun playing Half-Life 2, Homeworld 2, NWN, Pirates, Empire Earth II, Battlefield Vietnam, etc. etc. etc. on your PS-2.
Not to mention broadband gaming, which sucks royal ass on consoles.
don’t forget the Anti-Mac Cabal who, rather than look at the facts come out and claim the inferiority of the G5 even though the G5 has no bearing on the performance of Doom 3 after all the logic is taken care of.
you are really something.
Go have fun with your 15 year old friends there pal.
Those who actually know the facts behind the technology will go on not caring what you have to say.
What do these timedemo benchmarks mean? I want to know if I can run Doom 3 on an iMac G5 smoothly and with good quality.
And: does Doom 3 support the widescreen monitors.
it does support widescreen modes, but you have to add them to the config file. you will need the 1.8GHz iMac minimum, since it is very cpu intensive. but im not quite sure if the rather weak geforce fx 5200 is fast enough. better wait for the mac demo and test it before you buy.
//you are really something.//
Thanks!
//Go have fun with your 15 year old friends there pal.//
Only 15-year-old (notice use of hypens, youngster) friends I have are my students.
//Those who actually know the facts behind the technology will go on not caring what you have to say//
These “facts” that you mention — what are they again? I didn’t see them in your post.
Awaiting your intelligent reply …
I was truly disappointed by these results. The fact that the test machine was the fastest mac that money can buy with loads of RAM and a top-end graphics card makes these mediocre frame rates even harder to stomach.
You get the usual zealot response, blaming every non-Apple link in the chain: ATI drivers, compiler optimisation, etc, etc. At the end of the day, does it really matter whose fault it is? The bottom line is that the Doom 3 experience on the fastest mac hardware falls well short of the mark. That’s a real shame.
Well, keep in mind it’s a SINGLE game and was PORTED.
A “real shame”? Not really.
D3 sucks, regardless of how fast it runs.
“Wow. These benchmarks are also all showing the performance of PCs with PCIe graphics cards. Obviously, because of the greater bus throughput, the latest games are going to show better performance on a PC with PCIe than a Mac which currently only has AGP”
No. All benchmarks on PC comparing AGP systems to their PCIe equivalents show that this has no impact whatsoever – the performance is exactly the same.
//Your opinions are very immature, and not because they are counter to mine, but because your arguments are irrational and infantile at their roots.//
Being the erudite debater that you are … please specify which arguments you’re referring to, and explain their irrationality.
That’s the best reply you can give him? I was was hoping for some backup to your comments. How disappointing.
These types of benchmarks are stupid. Top max frames per second? Big deal! It’s not like the human eye is going to see a difference.
The real test is to put the system under extreme load and see if it slows down to the point that the human eye will notice.
If it doesn’t drop below 24 frames per second you’re fine.
http://en.wikipedia.org/wiki/Persistence_of_vision
What people really should be gripping about is how Mac users get ripped off when purchasing these conversion. Sure, we get the game, but we don’t get the editing tools like our PC using bretheren. Most cases we pay the same or more for a game that is incomplete!
well, i was getting <20fps. i changed one number in a config (something to do with graphics card apurture being set by default to 32megs or something retarded like that), and i was getting ~50. i am also far from the only one who experienced this.
Perhaps *you* know nothing about technology. Or what you think you know is half-assed tripe. Probably the latter.
really? you *have* been reading these comments, havnt you?
“the architecture of the g5 HAS NOTHING to do with the performance of this game !!! i know many of u are hoping that this proves that intel or amd chips are better than the G5 but that is not the case !!!”
of course. cpu archetecture has NO BEARING on game performance. right. and game performance dictates the quality of a cpu. uh huh.
“Like most apps, it goes slower on Mac than compared to Windows. Mac is not optimized for speed. Not at any level.
Nor is Mac optimized for memory usage. The same app on Mac (say Firefox) takes much more memory on Mac vs. Windows. And will run much slower at the same Mhz. ”
firefox uses more memory on osx, that means osx isnt optimized. mhz has everything to do with processing speed, and ppc/x86 comparisons are judgable in mhz. really? and no piece of bsd has been compiled with optimization flags? i can believe that.
“G5 is faster in interface response but the graphics processes are still slower than PCs. I think OSX loses performance in some place that Apple cannot change and the company is waiting for bigger and faster processors to hide this issue.”
yes, because a geforce fx gpu on a mac card is slower then the same on a x86 card. osx losses performance in areas that apple, the people who make both the hardware and the software, cannot change. really. and they are just sitting around, waiting for bigger processors, even though with every version of osx, everything gets substancially faster.
this thread is full of uninformed people who think they know what they are talking about. i have found this normal amoung gamers, who know what video card is better then the other (when they arnt caught up in ati/nvidia religious zealotry), and thats about it.
I thought they were pretty self evident.
Rockwell said:
“Have fun playing Half-Life 2, Homeworld 2, NWN, Pirates, Empire Earth II, Battlefield Vietnam, etc. etc. etc. on your PS-2.
Not to mention broadband gaming, which sucks royal ass on consoles.”
hmm… yeah that sure sounds like an adult who teaches high school.
Rockwell also said:
“Funny, I tweaked no config files and got 48-58 FPS throughout, on my Athlon 2800+ with GeForce FX 5700 stock.
Perhaps *you* know nothing about technology. Or what you think you know is half-assed tripe. Probably the latter.”
To me, you sound like smart 18 year old kid who has a lot of growing up to do.
//If it doesn’t drop below 24 frames per second you’re fine. //
I call bullshit. I can _immediately_ notice a difference between 60 fps and even 40 fps when playing a first-person shooter. If you looked closely, you would too. This probably comes from years of playing these kinds of games, but it *is* noticeable.
And, for what it’s worth — the G5 is a fine computer, and I have no issues with them (and, to be honest, sometimes I wish I could afford one myself — their design is kick-ass).
But, for gaming, I steadfastly argue that WinTel will give you much better performance for the price. At least for now. Perhaps that will change in the future, but in the meantime it’s something you just gotta live with, if you go the Mac route.
You gave me another example!!!!
Rockwell said:
“I call bullshit. I can _immediately_ notice a difference between 60 fps and even 40 fps when playing a first-person shooter. If you looked closely, you would too. This probably comes from years of playing these kinds of games, but it *is* noticeable.
And, for what it’s worth — the G5 is a fine computer, and I have no issues with them (and, to be honest, sometimes I wish I could afford one myself — their design is kick-ass).
But, for gaming, I steadfastly argue that WinTel will give you much better performance for the price. At least for now. Perhaps that will change in the future, but in the meantime it’s something you just gotta live with, if you go the Mac route.”
please keep them coming.
//I thought they were pretty self evident.//
Er … come again? Those weren’t arguments, just facts about gaming.
You blathered on like a dipshit about how my “arguments are irrational and infantile at their roots” … and I wasn’t even arguing anything.
Fact: You cannot play Half-Life 2, Homeworld 2, NWN, Pirates, Empire Earth II, and Battlefield Vietnam on the PS/2 (at least, not yet).
Fact: I tweaked no config files and got 48-58 FPS throughout Doom 3, on my Athlon 2800+ with GeForce FX 5700 stock.
So … again, please explain how these are “irrational and infantile” arguements.
Your utter stupidity is really starting to show through loud and clear … please continue to embarass yourself …
I think what it comes down to is that if you are buying a Powermac just to play games, unless you will be playing Quake 3 exclusively, you will be sorely disappointed.
Disclaimer: My primary computer is a Powermac G5 Dually and I have a PC for the occasional game.
“An article containing SOME substance being published on MacWorld. I can’t believe it. Usually they are merely a parrot for Apple.”
Actually the past few articles have been really good and they have been sticking it to Apple.
With speed increases once every year as it’s going now. Apple will loose more of it’s pro-user base.
Lets face it, games are to expensive to play on a dual G5, when anyone can get a gaming machine for a grand on the Wintel side. One can build one easily for that and all the games are optimised for Windows.
thats assuming pro-user is the same thing as gamer. considering the state of mac gaming since the 80s, i doubt it will have much of an impact.
//you know just as I do that in the context presented they were arguments. you were putting up a counter example to a statement made by another. that is an argument no matter how you spin it.//
Uh … ok, whatever. I can’t even think of a way to respond to you anymore. I’ll just let your comments speak for themselves.
But, for gaming, I steadfastly argue that WinTel will give you much better performance for the price. At least for now. Perhaps that will change in the future, but in the meantime it’s something you just gotta live with, if you go the Mac route.
agreed. this is the way it has alwas been, stuff has gotten a bit better but not by much. personally, i play world of warcraft and enemy-territory, and thats about it. theres rarely more then a few good games people are playing anyways, and because of factors like a job, and a girlfriend, and a life, i only play two of them (which is already too much, damn blizzard and their perfect games). when i was a hardcore gamer, it was pc all the way. best feature of windows is that its users have no ethics, so software/games are readily available if you know where to look. MUCH harder to find warez on a mac. and the number of game publishers who put out crap is staggering (doom3 is the perfect example of a god-awful game that got hyped to death and did well in sales) if you are a gamer, get a modded console or windows, because the only way to know if a game is or isnt a waste of money is to start playing. if you are a gamer with ethics, buy games that you like and feel are worth the money. and if you are either, stay the hell away from the mac, its awsome for work, not as awsome for play (although, i do find it makes work more fun, but thats a totally subjective opinion and i dont want to start a(nother) flame war)
Why has nobody mentioned the difference between OpenGL and DirectX driver performance on the PC? In general, both Nvidia and ATI’s DirectX drivers give better performance than the OpenGL version. Macs don’t have the option of DirectX. Over time, I’m sure the drivers will improve as always.
While Aspyr may have tried to optimize for Mac, fundamentally the game was designed to run well on a PC. This will give it an inherent leg up in any comparison regardless of how hard Aspyr works at it.
I figure that the PC will always have more cutting-edge games and performance, but that doesn’t mean that games aren’t playable on the mac.
//and if you are either, stay the hell away from the mac, its awsome for work, not as awsome for play (although, i do find it makes work more fun, but thats a totally subjective opinion and i dont want to start a(nother) flame war)//
Also agreed. I might always have a PC, just for gaming. However … I keep eyeing that Mac Mini … heck, for $500 … it might be fun to get one, just to see what all fuss is about. You say it makes work more fun? I’m interested in your opinion about that … how so?
Usually when some one is left speechless, it means thy cannot argue their point anymore. I am sure you are the exception however. I would never fear my comments standing on their own, if I did I would not post on the web.
Mac mini is not “more fun”. It has the limitation of 1GB of RAM and 32MB VRAM and if you are coming from Windows and are used to a certain snappiness, it is not to be found on a low-end Mac.
Like the Anandtech review of the PowerMac stated, it takes at least a Dual 2.5Ghz with the Ultra 6800 video card to achieve the same level of crispness as a basic Windows machine.
We have one in the labs and it was money down the drain.
It will run out of memory extremely quickly if you use Safari or OmniWeb (or anything that uses Apple’s WebCore). This is because Apple’s WebCore appears to be written in Java or have an extremely flawed memory architecture. Web pages takes 10X or more the memory they do on Windows. With either of these two browsers, you can open only 40-60 tabs and then your machine will hit time stop.
And contrary to public mythology, Apple’s virtual memory system basically does not work. Once you are out of physical RAM, your machine slows to a crawl. It is not like Windows which has is well optimized and works quickly.
Overall, Mac has a nice aesthetic. But the performance does not compare to Windows and the reliability of the operating system is not even at Windows XP level. Apple issues patches like Microsoft which break many apps. The testing appears to be minimal before patches are unleashed on the public.
Apple does not tell you this when you buy a Mac mini — 32MB VRAM will not work well with OS X 10.4 Tiger. So you are buying a deadend computer that will never adequately run Apple’s latest operating systems.
usability is trendy now, but it wasnt alwas that way. apple has been built from day one with the idea that you design the interface around the strengths and limitation of the human brain, rather then around the way the machine or task works. if your at all interested in design, im about half way through jef raskins “The Humane Interface”, and would consider it required reading for anyone who wants to understand how interfaces work, and how one can be measurably better then another.
because of the macs design, it literally takes less brain power to use. it is far from perfect, but it is significantly better then anything else around from a design standpoint. there was a quite flamey style article on here recently about some programmer chick who hated all guis till she went to mac. at one point she says that using a windows machine leaves her far more exhausted then using a windows box. the reason for this is that there are all sorts of mental hoops that you jump through without being aware of every time you use a computer, and it is significantly less on a mac.
if you do pick up a mac mini to try it out, i can almost guarentee you wont like it too much at first, at the least you will not be anywhere near “comfortable” with it until you adjust to a totally different usability paradigm. when you do though, you will see what im talking about and finally understand why there are so many mac zealots out there. its rare that you get a product thats fun to use, and apple is one of the few companies that consistantly turn them out (if you want another example, the experience of using an ipod is completely different from any other mp3 player ive used. features is about half the reason i like mine so much at any rate) mac zealots tend to be people who know next to nothing about technology, and rarely even understand the real reasons that they like their macs so much. (which is why these mac vs pc arguments get so emotional, it stops being about the os cause neither party knows what its strengths are)
thanks for the excellent example of why mac and pc owners speak different languages.
Mac mini is not “more fun”. It has the limitation of 1GB of RAM and 32MB VRAM and if you are coming from Windows and are used to a certain snappiness, it is not to be found on a low-end Mac.
Like the Anandtech review of the PowerMac stated, it takes at least a Dual 2.5Ghz with the Ultra 6800 video card to achieve the same level of crispness as a basic Windows machine.
from what i have heard from users on similarily speccd machines is that pretty much everything is cool except expose on a 32meg card. you cannot compare windows and osx crispness, that is like comparing windows to beos. the technology is wayyyyy more advanced on the mac, of course it needs better hardware.
We have one in the labs and it was money down the drain.
It will run out of memory extremely quickly if you use Safari or OmniWeb (or anything that uses Apple’s WebCore). This is because Apple’s WebCore appears to be written in Java or have an extremely flawed memory architecture. Web pages takes 10X or more the memory they do on Windows. With either of these two browsers, you can open only 40-60 tabs and then your machine will hit time stop.
apple uses KHTML, which is developed for and by the kde guys. if you use ie in windows, then you are far less intelligent then you seem to be, if not then i would like to point out that both firefox and camino exist for mac using the best technology currently available for browsers (best being best supported and implemented standards based rendoring). and why in the world would you have 60 tabs open?
And contrary to public mythology, Apple’s virtual memory system basically does not work. Once you are out of physical RAM, your machine slows to a crawl. It is not like Windows which has is well optimized and works quickly.
actually, the virtual memory does work, but the paging system isnt that great. if you throw about a gig at it, multi-tasking is much, much, much smoother then under windows, no matter how much hardware you throw at it.
Overall, Mac has a nice aesthetic. But the performance does not compare to Windows and the reliability of the operating system is not even at Windows XP level. Apple issues patches like Microsoft which break many apps. The testing appears to be minimal before patches are unleashed on the public.
Apple does not tell you this when you buy a Mac mini — 32MB VRAM will not work well with OS X 10.4 Tiger. So you are buying a deadend computer that will never adequately run Apple’s latest operating systems.
here we go, the reason i am responding. there is alot more to using a computer then the specs. it doesnt matter if the app runs at half again the speed on windows if it takes you twice as long to get work done. i will almost guarentee that you wouldnt realise it, even if you had gone through the adjustment period i mentioned up above somewhere. human perception of time is totally dependant on what your brain is doing. a good example is that it is a common belief that keyboard text selection is faster then a mouse. turns out that a mouse is almost twice as fast, and pretty much to the man everyone will say the keyboard is faster. this was proven through some studies at apple, and the tog talks about it in one of his articles (too lazy to dig up the link, but his site is http://www.asktog.com). you dont know what makes you more productive. most developers have as little clue and make horrible interfaces. companies like microsoft started out bad, and have steadily been getting better (you see fittes law being leveraged more and more with every version of the os), but they are bound by the fact that a major change would make all their current users feel totally out of place, which would mean less sales, and of course, cash is what matters. the mac on the other hand has been designed to work for you since day one, and while osx has introduced alot more stupidness then was previously there, the tradeoffs seem to have been worth it.
Why has nobody mentioned the difference between OpenGL and DirectX driver performance on the PC? In general, both Nvidia and ATI’s DirectX drivers give better performance than the OpenGL version. Macs don’t have the option of DirectX. Over time, I’m sure the drivers will improve as always.
You are wrong here on 2 things. It is common knowledge that ATI drivers are better with handling the DirectX API than OpenGL and NVIDIA is better with OpenGL (D3 runs faster on NV cards on the PC).
And there is a wrapper for DirectX on the Mac called MacDX [http://www.coderus.com] but I haven’t heard of any game that would use it.
While Aspyr may have tried to optimize for Mac, fundamentally the game was designed to run well on a PC. This will give it an inherent leg up in any comparison regardless of how hard Aspyr works at it.
The game was designed to run on a wide range of platforms. They wouldn’t have used OpenGL if porting wasn’t their first pripority. The article also mentiones that they worked close with Apple to optimize it for the Mac.
I figure that the PC will always have more cutting-edge games and performance, but that doesn’t mean that games aren’t playable on the mac.
Depends on what game you’re talking about. Someone mentioned earlier that Quake 3 ran faster on the Mac as on the PC. No one (at least, not me) is talking that Mac is not a good platform for games, because it obiously is, but this I consider a major letdown because id’s technology is one of the very rare that still uses a portable API and is widely accepted. Most of games running id’s technology have been ported to the Mac, but as it seems (right now) the Doom 3 engine doesn’t work well with it, so you can expect future games using it’s technology to be very slow on Mac (that is, if they choose to port them) as they will undoubtly push the [the Doom 3] technology further.
Remember that people buy new systems to play id games. Imagine what kind of an impact on the Mac would it be, if D3 would run (let’s just pretend) twice as fast on the same hardware (3D card that is).
If the Xbox version runs smoother than on a dual 2.5GHz G5 w/ an x800 or 6800, then somewhere something is terribly wrong. *sigh*
YOu should research what OS X does with memory and what Windows does with memory.
OS X attempts to maximize the use of memory available while windows attempts to minimize the use of memory available. what use is memory if you are not using it if you add more?
full memory usage doe snot mean bloat.
It would make it easier for people to read what is going on and readers could see the argument as it goes on or totally skip the argument and go to the next one. Heck, you could default to hide the thread so people can just quickly look over the topics being discussed and see if there is anything of interest with out having to wade through the junk.
With either of these two browsers, you can open only 40-60 tabs and then your machine will hit time stop.
You should change your usage behaviour. Since you’re coming from windows, a little tip: the round thingy on the tab with the cross on it is for closing the tab.
And contrary to public mythology, Apple’s virtual memory system basically does not work. Once you are out of physical RAM, your machine slows to a crawl. It is not like Windows which has is well optimized and works quickly.
Pardon? I didn’t know Windows can do magical things. If an OS with VM runs out of memory, it has to use the hard disk which generally means performance penalty of a factor of 10^3. Since I work with PCs but use a much lower spec’ed Mac at home, I can assure you that there is absolutely nothing wrong with the VM system of OSX. As you observed yourself, OSX can handle memory hungry apps quit well. If I try to validate an XML file (around 12 MB size) on the PC with 1GB RAM/P4 1.7GHz, I can go for lunch because Windows get’s unresponsive for at least half an hour thanks to it’s ‘superiour’ VM. Laughable…
But the performance does not compare to Windows and the reliability of the operating system is not even at Windows XP level.
How the hell can you compare the “performance” of a system? Maybe you can compare certain functionality e. g. file copy operations where OSX is BTW years ahead of MS in performance and reliability (having lost more than once files on the file server because of Windows Explorer bugs). That does not mean that OS X is faster in general nor can you claim the same for Windows.
Apple issues patches like Microsoft which break many apps.
I’ve gone from 10.1 to 10.3.8 and do not agree with your statement. And in contrast to MS “patches” (= security fixes), most OS X “patches” are OS upgrades.
32MB VRAM will not work well with OS X 10.4
Huh? How do you know? I use a G3 B/W with 8MB VRAM which does not support Quarz Extreme. However that does not mean that you can’t work “well” with OSX 10.3 and I’m quite sure that this setting will be fast enough for me to use with 10.4. Your “work well” term depends on the intendet usage. And I there’s more than way where Windows does not “work well”.
…is in the arena.
Is there a multi-player mode?
The thing is that directx usually has more features, or at least usually has them first, opengl usually lags behind and then provides this features via wierd extensions to open (like nv_ extensions etc).
On my nvidia ti4200 doom3 is slower and looks like crap compared to half life 2 or farcry (both directx), but that’s just me.
OpenGL vs DirectX:
OpenGL is an industry standard and has a standards board; this is why most games systems support it. DirectX only exists in Windows. The Mac’s OpenGL system is faster then the Windows OpenGL system. A DirectX game ported to a Mac will be slower since the port normally just introduces a DirectX to OpenGL conversion layer.
The Benchmarks (Where to start…):
1) D3 doesn’t use multi-processors; thus the benchmark comparisons are actually a 2.5Ghz G5 is only a little slower then a 3.4Ghz P4. If Apple made a single processor 2.5GHz box, the benchmark numbers would be the same.
2) Comparing famerates is basically stupid; when you get a boss screen and the rates go way down, it’s normally the game getting in the way more then the screen rendering. For those of you who think you can tell the difference between 50 and 60FPS, please compare it watching a move (TV or Theatre); please note that in both cases the screens show at only 24FPS. There are only a few people (less then about 1% of the population) that can’t watch TV because frame rates are too slow and they can actually see the black bars between the frames. Anything over 30FPS is a waste. Everyone, talking about game consoles being better for gaming then PC’s; just remember that you’re only getting 24FPS from the console.
Only problem is that D3 is OpenGL NOT DirectX. D3 was written from the ground up using OpenGL. The issue is OSX OpenGL and posssibly the video drivers. As it is optimized more and more the speed will come but its not there yet, as we see.
After reading other Q&A from the tester (Peter Cohen) the actual ‘in game’ FPS counter was used and it rarely dropped under 40 @ 1024×768 and 1280×1024.
http://www.insidemacgames.com/forum/index.php?showtopic=19287
Also there are comments directly from Glenda Adams of Aspyr who is in charge of the port.
> you’re only getting 24FPS from the console.
Not true. All modern consoles output 60 fps on NTSC and 50 fps on PAL. (or 30 and 25, depending on the game)
Some Altivec specific code wouldn’t hurt. I couldn’t care less about Doom3, it’s one of the most boring and linear games ever, but if the engine gets ported this could lead to better gaming on the Mac, not just silly FPS games but games where you have to think and that sport excellent 3D graphics.
This is a typical example of not using the right tool for the job, IBM chip so use the damn IBM compiler ID you muppets!!
Maybe they should read this- http://homepage.mac.com/WebObjects/FileSharing.woa/wa/downloadFile?…
Personally, I think the performance hit is down to OSX’s OpenGL implentation and the drivers for the ATI & Nvidia cards. I use my ibookG4 for everything but games – those I use my Shuttle XPC Athlon 2500+ Ati 9800Pro.
Anon
“I oft find it amazing how a OS known for not multitasking well also does not single task well either”
What the hell are you talking about? *nix systems have far better multitasking than Windows does. In windows, if a program decides to go errant and take 100% CPU time, you can’t even get the task manager up. You are forced to hard reboot. Even if you manually tell the program to have lowest priority before you run it, it still does that. I can’t use single CPU windows boxes for that reason (yeah, I crash things a lot). You don’t run into that on any *nix system that is set up right.
*nix had better multitasking than Windows does 30 years ago.
I wonder if they tried changing the priority of the game? Something like “nice -20 doom3” would have improved performance, I’m sure. Maybe drastically, maybe not, but it definitely would be better. That would give you your “single tasking” pretty well.
Not true. All modern consoles output 60 fps on NTSC and 50 fps on PAL. (or 30 and 25, depending on the game)
isn’t that 60 INTERLACED Frames? Which in essence is 30 frames.