Quake III Arena Benchmark Thread

Discussion about old graphics cards, monitors and video related things.

Re: Quake III Arena Benchmark Thread

Postby dexvx » 2017-8-13 @ 17:48

Ok... is the Radeon 9700 Pro supposed to be slower than a GeForce 4 Ti4600? I always remember the Radeon 9700 Pro as being WAAAY better than the GeForce 4's.

Here's my result (same system as my posted Ti4600 results).

242.3 fps, dexvx, Hercules Radeon 9700 Pro 325/620 (stock, Cat 10.2), P4 3.06/133 (stock, 1C/1T), Asus P4S533 (SiS 645DX), 1GB DDR 333

Edit: Hmph, looks like it's about on par (4% faster) in Q3A at 1024x768x32. Wonder why mine is slower. Maybe SiS P4 chipset?

http://www.anandtech.com/show/947/22
dexvx
Oldbie
 
Posts: 725
Joined: 2017-3-07 @ 03:32
Location: USA

Re: Quake III Arena Benchmark Thread

Postby slivercr » 2017-8-14 @ 01:01

dexvx wrote:Ok... is the Radeon 9700 Pro supposed to be slower than a GeForce 4 Ti4600? I always remember the Radeon 9700 Pro as being WAAAY better than the GeForce 4's.

Here's my result (same system as my posted Ti4600 results).

242.3 fps, dexvx, Hercules Radeon 9700 Pro 325/620 (stock, Cat 10.2), P4 3.06/133 (stock, 1C/1T), Asus P4S533 (SiS 645DX), 1GB DDR 333

Edit: Hmph, looks like it's about on par (4% faster) in Q3A at 1024x768x32. Wonder why mine is slower. Maybe SiS P4 chipset?

http://www.anandtech.com/show/947/22


In the q3a SMP scaling thread I get "parity" between a 9800 Pro and a Ti4600 when SMP is disabled, and the Ti4600 is faster with SMP enabled. Seems ok that the 9700 Pro is a tad slower.

No clue as to why, though.
Outrigger: an ongoing adventure with the OR840
QuForce FX 5800: turn your Quadro into a GeForce
User avatar
slivercr
Member
 
Posts: 307
Joined: 2017-2-16 @ 18:00
Location: Groningen

Re: Quake III Arena Benchmark Thread

Postby appiah4 » 2017-8-14 @ 09:14

ATI OpenGL drivers most likely. There is also the fact that the Q3 engine greatly predates the Radeon 9700 and DirectX9, it was built using OpenGL1.x which was probably not something ATI particularly bothered optimizing for in 2002. Also, that pesky P4 is basically cpu bottlenecking the 9700 which can easily scale further with an Athlon64 whereas the Ti will likely not.
A500:Rev6|+512K|ACA500+|C1084S
i386:Am386SX25|4M|GD5402|ES688
i486:U5S33|8M|GD5428|YMF719|DB-S2
i586:P133|32M|T64+/MX2|V1|CT3980/32M
i686:K6-2/400|128M|Rage|V2|CT4520/32M
S370:P3-1200|384M|GF4-4200|MX300
S754:A3700+|2G|X1950PRO|SB0350
User avatar
appiah4
l33t
 
Posts: 4172
Joined: 2017-2-19 @ 07:36

Re: Quake III Arena Benchmark Thread

Postby dexvx » 2017-8-14 @ 17:01

I'm blaming the SiS chipset for now, I know the 645DX is quite a bit slower than the more expensive i850/E. I've seen some benchmarks where the i850E it is 20% faster with PC1066 RDRAM. Unfortunately, my TH7II bios is corrupt, so I have no way of using my only i850. Need to find someone with a bios flasher locally.
dexvx
Oldbie
 
Posts: 725
Joined: 2017-3-07 @ 03:32
Location: USA

Re: Quake III Arena Benchmark Thread

Postby slivercr » 2017-8-14 @ 20:54

dexvx wrote:I'm blaming the SiS chipset for now, I know the 645DX is quite a bit slower than the more expensive i850/E. I've seen some benchmarks where the i850E it is 20% faster with PC1066 RDRAM. Unfortunately, my TH7II bios is corrupt, so I have no way of using my only i850. Need to find someone with a bios flasher locally.


I dont think its the chipset being "slow", since I have similar results with an i840 based motherboard: your PC is a faster platform than mine, it just moved the maximum FPS up but kept the trend of the ti4600 being a bit faster.

I think it has to do more with the drivers you are using and the resolution being tested. The trend will persist until the Ti4600 cannot handle the resolution, if you test at 1080p the 9700 Pro will be faster by far, for example.
Outrigger: an ongoing adventure with the OR840
QuForce FX 5800: turn your Quadro into a GeForce
User avatar
slivercr
Member
 
Posts: 307
Joined: 2017-2-16 @ 18:00
Location: Groningen

Re: Quake III Arena Benchmark Thread

Postby leileilol » 2017-8-14 @ 21:30

*cough*/r_primitives 2*cough*

A lot of the judgment against ATI had to do with the compiled vertex array extension not being properly detected by the game on their cards making it pick a slow path (r_primitives 1 uses this fallback mode).
by the way, DOSBox is not for running Windows 9x
User avatar
leileilol
l33t++
 
Posts: 9826
Joined: 2006-12-16 @ 18:03

Re: Quake III Arena Benchmark Thread

Postby slivercr » 2017-8-14 @ 22:03

leileilol wrote:*cough*/r_primitives 2*cough*

A lot of the judgment against ATI had to do with the compiled vertex array extension not being properly detected by the game on their cards making it pick a slow path (r_primitives 1 uses this fallback mode).


Ahh, cool! I'll play with this over the weekend.

You are very knowledgeable of these things, are there any other card specific console settings one should keep in mind?
Outrigger: an ongoing adventure with the OR840
QuForce FX 5800: turn your Quadro into a GeForce
User avatar
slivercr
Member
 
Posts: 307
Joined: 2017-2-16 @ 18:00
Location: Groningen

Re: Quake III Arena Benchmark Thread

Postby leileilol » 2017-8-14 @ 22:56

Probably r_ext_compress_textures. That defaults to 1 as S3 cards need it and id didn't foresee any other card adopting S3TC at the time. It can be safely set to 0 on 32mb+ cards

The rest of the renderer cvars only really matter for earlier video card generations (and are often misled on many "q3 tweak config" pages)
by the way, DOSBox is not for running Windows 9x
User avatar
leileilol
l33t++
 
Posts: 9826
Joined: 2006-12-16 @ 18:03

Re: Quake III Arena Benchmark Thread

Postby Fusion » 2017-8-15 @ 05:17

I agree with leileilol, those tweak guides were poison to my framerates back in the day all because I didn't know any better. :( There are very few q3config tweaks that do much at all.
WIP Retro Files
P3 800 | 512MB PC100| PCI V3 3000 16MB @ 195/195 | CT4780 SB Live! Value| WinME
P3 450 | 128MB PC100| AGP TNT2 Pro 32MB | CT4170 SB16 | MS-DOS 7.10
A64 2.4 | 2GB DDR1 | PCI-E Radeon X800XT @ 575/575 | Win2K
User avatar
Fusion
Member
 
Posts: 192
Joined: 2017-3-06 @ 04:49
Location: Ontario, Canada

Re: Quake III Arena Benchmark Thread

Postby Ponjiayulady » 2017-8-18 @ 10:41

I also test quake iii with my old stuffs
I'm happy to get my old p3 1Ghz slot1 working again :happy:

my components:
pentium iii 1Ghz slot1
256MB PC133 SDRAM
Voodoo3 2000 and TNT2 mach64
Win2K system
1g use3.JPG


here is my result:
Voodoo3 2000 AGP
1g use2.JPG
1g use4.JPG

Riva TNT2 mach64
1g use1.JPG
1g use5.JPG
Ponjiayulady
Newbie
 
Posts: 23
Joined: 2017-8-18 @ 09:41

Re: Quake III Arena Benchmark Thread

Postby silikone » 2017-8-18 @ 14:00

How come Quake 3 benefits so much from a Geforce 3 over a Geforce 2 compared to other titles?
Do not refrain from refusing to stop hindering yourself from the opposite of watching nothing other than that which is by no means porn.
User avatar
silikone
Member
 
Posts: 259
Joined: 2012-3-21 @ 19:53

Re: Quake III Arena Benchmark Thread

Postby appiah4 » 2017-8-18 @ 14:17

silikone wrote:How come Quake 3 benefits so much from a Geforce 3 over a Geforce 2 compared to other titles?


It's built around the idea of using programmable shaders, something GF2 lacks.
A500:Rev6|+512K|ACA500+|C1084S
i386:Am386SX25|4M|GD5402|ES688
i486:U5S33|8M|GD5428|YMF719|DB-S2
i586:P133|32M|T64+/MX2|V1|CT3980/32M
i686:K6-2/400|128M|Rage|V2|CT4520/32M
S370:P3-1200|384M|GF4-4200|MX300
S754:A3700+|2G|X1950PRO|SB0350
User avatar
appiah4
l33t
 
Posts: 4172
Joined: 2017-2-19 @ 07:36

Re: Quake III Arena Benchmark Thread

Postby lazibayer » 2017-8-19 @ 22:42

Just for laughs...
2.7 FPS, Rage Pro AGP 2x 8MB, K6-3+ 450 OC 133x4.5=600, 256MB PC133 2-2-2, P5A-B, Windows XP, stock driver, 1024x768, 16bit color, 32bit texture.
17.8 FPS, Rage Fury (128GL) 32MB, K6-3+ 450 OC 133x4.5=600, 256MB PC133 2-2-2, P5A-B, Windows XP, driver version 6.13.3279, 1024x768, 32bit color, 32bit texture.
26.5 FPS, FireGL2 64MB @PCI mode, K6-3+ 450 OC 133x4.5=600, 256MB PC133 2-2-2, P5A-B, Windows XP, driver version 6.12.10.2106, 1024x768, 32bit color, 32bit texture.
Last edited by lazibayer on 2017-8-27 @ 22:58, edited 3 times in total.
User avatar
lazibayer
Oldbie
 
Posts: 1064
Joined: 2014-10-21 @ 21:16

Re: Quake III Arena Benchmark Thread

Postby dexvx » 2017-8-20 @ 01:16

appiah4 wrote:
silikone wrote:How come Quake 3 benefits so much from a Geforce 3 over a Geforce 2 compared to other titles?


It's built around the idea of using programmable shaders, something GF2 lacks.


Also at 1024x768x32, we start to see the effects of GF3's LMA (reducing draws/z-buffer compression).
dexvx
Oldbie
 
Posts: 725
Joined: 2017-3-07 @ 03:32
Location: USA

Re: Quake III Arena Benchmark Thread

Postby silikone » 2017-8-22 @ 20:45

appiah4 wrote:
silikone wrote:How come Quake 3 benefits so much from a Geforce 3 over a Geforce 2 compared to other titles?


It's built around the idea of using programmable shaders, something GF2 lacks.


I'm ignorant on this front, but I thought these shaders in the game were all software, separate from the programs that the NV20 can run.
Do not refrain from refusing to stop hindering yourself from the opposite of watching nothing other than that which is by no means porn.
User avatar
silikone
Member
 
Posts: 259
Joined: 2012-3-21 @ 19:53

Re: Quake III Arena Benchmark Thread

Postby spiroyster » 2017-8-22 @ 21:42

silikone wrote:
appiah4 wrote:
silikone wrote:How come Quake 3 benefits so much from a Geforce 3 over a Geforce 2 compared to other titles?


It's built around the idea of using programmable shaders, something GF2 lacks.


I'm ignorant on this front, but I thought these shaders in the game were all software, separate from the programs that the NV20 can run.

Yeah, the're not fragment shaders. It's a homegrown 'shader' system (mainly glorified texture combiners/multi-stage). so I would have thought any gains would come from increased texture performance and not inclusion of hardware accelerated programmable shaders.
User avatar
spiroyster
Oldbie
 
Posts: 538
Joined: 2015-10-12 @ 12:26

Re: Quake III Arena Benchmark Thread

Postby dexvx » 2017-8-24 @ 05:08

leileilol wrote:*cough*/r_primitives 2*cough*

A lot of the judgment against ATI had to do with the compiled vertex array extension not being properly detected by the game on their cards making it pick a slow path (r_primitives 1 uses this fallback mode).

Probably r_ext_compress_textures. That defaults to 1 as S3 cards need it and id didn't foresee any other card adopting S3TC at the time. It can be safely set to 0 on 32mb+ cards


with \r_primitives 2\

245.2 fps

with 'r_ext_compress_textures 0' (with restart)

245.4 fps

with both 'r_primitives 2' and 'r_ext_compress_textures 0'

245.1 fps

I think I'm platform bottlenecked. Man, I really need to fix my TH7II.
dexvx
Oldbie
 
Posts: 725
Joined: 2017-3-07 @ 03:32
Location: USA

Re: Quake III Arena Benchmark Thread

Postby Fusion » 2017-8-24 @ 22:36

Beat my old score! :lol:

51.6 FPS, Fusion, Voodoo3 3000 16MB PCI @ 195/195, Pentium III 800EB, Intel 810e, 384MB PC133, WinMe
Last edited by Fusion on 2017-8-26 @ 15:50, edited 1 time in total.
WIP Retro Files
P3 800 | 512MB PC100| PCI V3 3000 16MB @ 195/195 | CT4780 SB Live! Value| WinME
P3 450 | 128MB PC100| AGP TNT2 Pro 32MB | CT4170 SB16 | MS-DOS 7.10
A64 2.4 | 2GB DDR1 | PCI-E Radeon X800XT @ 575/575 | Win2K
User avatar
Fusion
Member
 
Posts: 192
Joined: 2017-3-06 @ 04:49
Location: Ontario, Canada

Re: Quake III Arena Benchmark Thread

Postby SPBHM » 2017-8-26 @ 07:51

all stock apart from the e2140.
816.3 FPS, SPBHM, Radeon HD 5850, Core i5 2310, H61, 8GB DDR3 1333 DC, Win 10 x64
301.7 FPS, SPBHM, Radeon HD 4670 DDR2 (750/400) PCIe, Pentium E2140@2133MHz, VIA P4M900, 3GB DDR2 533 SC, Win 7 SP1 x86

114.1 FPS, SPBHM, Geforce 8400GS G98 PCI, Pentium 4 2.4B, 845PE, 512MB DDR 333, Win XP SP3
(if I OC the CPU to 3GHz it gets 121FPS, if I OC the VGA to 675/580, up from stcok 567/500 it gets 131FPS, interesting that the 9250 PCI is clearly faster)


40.0 FPS, SPBHM, Voodoo 4 4500 PCI, Pentium 3 750MHz, 440BX, 256MB PC100, Win 98SE
(using 16bit to compare with the V3 resulted in 59.1, both with default GL driver, Voodoo GL driver doesn't seem to work)
Last edited by SPBHM on 2017-8-26 @ 20:04, edited 1 time in total.
SPBHM
Oldbie
 
Posts: 587
Joined: 2012-10-26 @ 15:59

Re: Quake III Arena Benchmark Thread

Postby MordekaiZeyo » 2017-8-26 @ 18:39

147.7 FPS, MordekaiZeyo, Galaxy Geforce 6600 GT 128 MB AGP, Intel Pentium 3-S 1266 Mhz, ECS P6S5AT with SiS 635T Chipset, 512 MB DDR-RAM, Windows 2000 Pro SP4

Quake3BenchMordekaiZeyo.png

1024x768 in fullscreen. I was lazy to take screenshot so i did with my smartphone. lol

I get about the same FPS on ALL screen resolutions in Quake 3 Arena.

About 150.8 FPS on 640x480 and about 144.x FPS somewhat on 1920x1080... It's limited by CPU funnily.
Intel Pentium 3-S 1266Mhz | 512 MB DDR-400 @SD-RAM PC133 | Elitegroup P6S5AT | Galaxy Geforce 6600 GT AGP @4x AGP | Terratec Xfire 1024 PCI Soundcard with Dreamblaster X2
Windows 2000 Pro SP4
MordekaiZeyo
Newbie
 
Posts: 21
Joined: 2017-3-25 @ 13:44
Location: Germany

PreviousNext

Return to Video

Who is online

Users browsing this forum: mpe and 2 guests