Thread Rating:
  • 14 Vote(s) - 3.79 Average
  • 1
  • 2
  • 3
  • 4
  • 5
GPU Benchmark designed for PCSX2
#1
updated to post #567



Instant Movement to the Hottest Benchmark Results:
NTSC PS2 BIOS - Harsh Group (2048x2048)





Quick Start Guide
Do not read each rectangle dialog box with heading "my explanation Wrote:" they are only explanations for your curiosity not instructions at all.




Prologue
GSDX10 is the choice from graphic plugins
my explanation Wrote:As you know GSDX10 have less graphic bugs i.e. higher compatibility than GSDX9
Despite I find it in real life that typically GSDX10 is a little slower than GSDX9
Compatibility is more important than speed this time


We benchmark PS2 BIOS Screen and DQ8
my explanation Wrote:Choose the most GPU-intensive game screen to make your GPU bottleneck
in the meantime, game speed in FPS is representation of GPU Performance






Step 1: Overclock your CPU as max as possible if you can
my explanation Wrote:This is GPU benchmark, your GPU should be bottleneck, so that overclock GPU will give you more FPS
If CPU is bottleneck, overclock CPU will give you more FPS, overclock GPU won't give you more FPS
FPS is limited by either CPU or GPU depending on which one is the slowest
Overclock your CPU to not let it potentially relatively slower than your GPU
Also you may get extra 1 or 2 fps on your current gpu clockrate, after overclocking your cpu to max






Step 2: nullify factors that affect pcsx2 game speed when GPU is bottleneck in pcsx2 gameplay
Note: If your GPU is stronger than 8600GT, you may not see a difference of FPS after applying one or more of the three operations below.
  1. Change screen resolution to 1024x768
    my explanation Wrote:higher screen resolution cause lower FPS and vice versa
    we use same screen resolution, so that we have same influence on pcsx2 game speed
    the choice is 1024x768 to take care of CRT users worldwide and 1024x768 is decent for PS2 3D games
     
     
  2. maximize pcsx2 game window
    Note: this is extremely important. I can see full statistics shown in pcsx2 game window title in maximized window, especially in where shows the CPU percentage that is utilized to identify whether a benchmark result is valid or not
    my explanation Wrote:larger pcsx2 game window cause lower FPS and vice versa
    we use same pcsx2 game window size, so that we have same influence on pcsx2 game speed
    it's maximized pcsx2 game window size under screen resolution 1024x768
     
     
  3. switch to Windows Classic theme (see screenshot if you dunno how to do it)
    my explanation Wrote:the default Windows Aero Theme is occupying a little GPU power
    thus reduce pcsx2 game speed by 1 or 2 FPS




Step 3: Prepare PCSX2
  1. Pcsx2 and Plugins (the same as CPU Benchmark designed for PCSX2 based on FFX-2)
    Using Official beta 1888 and plugins on March 11th 2010. But the plugin package lacks DEV9null, you can extract a DEV9null from the installer of beta 1888 by 7-zip.
    my explanation Wrote:We all using same version of pcsx2, cuz pcsx2 game speed may vary in different revisions of pcsx2 that each one uses ought to be different.

    Before you ask, I tell you Why this benchmark persists with PCSX2 0.9.6 1888 as newer version of PCSX2 available:
    This benchmark was created on May 16 2010 when the latest version of PCSX2 is 0.9.6 1888

    This benchmark is aiming at CPU Performance under PCSX2, any version of PCSX2 is adequate, it's not necessary to use up-to-date version of PCSX2, despite new version of PCSX2 typically have more FPS.

    What the hell if we redo all existing benchmark results with the latest version of PCSX2 ONCE A NEW VERSION OF PCSX2 RELEASED!
     
     
  2. All pcsx2 settings are default except for those I mentioned below
    my explanation Wrote:you should know there are no speedhacks or any other tricks by default settings
     
     
  3. configure pcsx2 from beginning
    We all select SSE2's GSdx, to take care of AMD CPU Users.
    Nothing special to configure in the Configuration window.
    [Image: Pcsx2_Frist_Time_Config.png]
    my explanation Wrote:the difference among different instructions' GSdx is CPU-wise, which can be ignored as this time here's the GPU Benchmark
     
     
  4. General GSdx Settings
    1. Renderer: Direct3D10 (Hardware)
    2. Aspect Ratio (F6): Stretch
      [Image: GSdx_General.png]
      my explanation Wrote:Choosing Stretch doesn't affect speed, utilized to eliminate black bars for a nice looking
     
     
  5. General Lilypad Settings
    your custom settings, assign keys to buttons in Pad 1, etc.
     
     
  6. Disable Frame Limiting
    In pcsx2 main menu you click Config -> Cpu, set Frame Limiting to normal you suppose your Video Card has enough power to go above normal speed (60 FPS for NTSC and 50 FPS for PAL).
    [Image: Pcsx2_Frame_Limiting.png]
     
     
  7. Tip: If you mess up your pcsx2 settings, delete the inis folder in pcsx2's main directory, then you can configure pcsx2 from beginning, which ensures all settings are default.




Step 4: Benchmark PS2 BIOS Screen
Note: Normal PS2 BIOS speed from PAL models' PS2 is 50 FPS, while NTSC's 60 FPS.
  1. Introduction to Mild Group and Harsh Group
    1. Mild Group:
      NTSC PS2 BIOS - Mild Group (1024x1024) and PAL PS2 BIOS - Mild Group (1024x1024)
      Using low 3D quality for low end Video Card worse than 9600GT or so
      note: Do not do the Mild Group benchmark if your Video Card is better than 9600GT
      my explanation Wrote:typically your bottleneck is CPU if your Video Card is better than 9600GT
      this is not what expected that GPU should be bottleneck in this GPU benchmark
      Unless overclock an i5 2500K then your CPU won't be bottleneck in Mild Group benchmark
       
    2. Harsh Group:
      NTSC PS2 BIOS - Harsh Group (2048x2048) and PAL PS2 BIOS - Harsh Group (2048x2048)
      Using high 3D quality for high end Video Card
      note: Your computer will become extremely slow and almost hang if do the Harsh Group benchmark on low end Video Card. e.g. 8600GT.
     
     
  2. GSdx Settings for Mild Group and Harsh Group
    1. Mild Group:
      1. Set "Interlacing" to "Blend bff" in GSdx Settings to suppress the vibration of bios screen
        Note: Some Video Cards including NVIDIA and AMD report different FPS from "Blend bff" to "None" in the setting of Interlacing, I'll ignore your benchmark results if you set Interlacing to None
      2. Set "D3D internal res" to "1024x1024" in GSdx Settings
      [Image: GSdx_BIOS_1024.png]
       
       
    2. Harsh Group:
      1. Set "Interlacing" to "Blend bff" in GSdx Settings to suppress the vibration of bios screen
        Note: Some Video Cards including NVIDIA and AMD report different FPS from "Blend bff" to "None" in the setting of Interlacing, I'll ignore your benchmark results if you set Interlacing to None
      2. Set "D3D internal res" to "2048x2048" in GSdx Settings
      [Image: GSdx_BIOS_2048.png]
     
     
  3. the PS2 BIOS Screen for record your FPS is Clock Adjustment under System Configuration
    No speed changes if the display language of BIOS is different
    Don't forget to maximize the pcsx2 game window before taking screenshot
    [Image: BIOS_1024_Screen_4526_FPS.jpg]




Step 5: learn identify your benchmark result is valid or not
The percentage of CPU shown in pcsx2 game window title should be at least 90% to ensure GPU is bottleneck. Or below 90% indicates CPU is bottleneck and your benchmark result is invalid.

An example of valid benchmark result, behold the cpu percentage in pcsx2 game window title. It's 97% CPU.
[Image: DQ8_US_Screen_3657_FPS.jpg]






(Optional) Step 6: Benchmark DQ8 (SLUS 21207)
As above Step 4: Benchmark PS2 BIOS Screen, the DQ8 Benchmark also has Mild Group and Harsh Group:
DQ8 (SLUS 21207) - Mild Group (1024x1024) by set "D3D internal res" to "1024x1024" in GSdx Settings
DQ8 (SLUS 21207) - Harsh Group (2048x2048) by set "D3D internal res" to "2048x2048" in GSdx Settings
  • Screenshot: Starting a new game up until to a rat rushing into your pocket.
    Don't forget to maximize the pcsx2 game window before taking screenshot.
    [Image: th_DQ8_US_Screen_3657_FPS.jpg]
  • or load the savestate below, bring your to the game screen above.
    F4715852.000






Step 7: Submit your benchmark results
The formation of a benchmark result:
Code:
FPS - Name of Video Card - Your user name at this forum - GPU-Z
  1. FPS: Full fps shown in pcsx2 game window title, e.g. 75.64 FPS
     
  2. Name of Video Card: Go check GPU-Z for the info if you're not sure. You should add the Manufacturer name of GPU(NVIDIA, AMD, Intel for Intel IGP including the one built into Core i3 or Core i5), e.g. NVIDIA GTX 580, AMD Radeon HD 6970.

    (additional information if necessary):
    e.g. NVIDIA GT240 (GDDR5), NVIDIA GT240 (GDDR3)

    (OC): represent your Video Card is overclocked.
    e.g. AMD Radeon HD 6950 (OC 1000/1425) where 1000 and 1425 stand for GPU Clock and Video Card Memory Clock.

    (Stock): represent your Video Card is not overclocked.
    e.g. NVIDIA GTX 480 (Stock 700/924)
    e.g. AMD Radeon HD 5870M. where the user doesn't specify (Stock) or (OC), I assume it's (Stock)

    You should at least post your result at Stock speed:
    Quote:First off pay attention to factory pre-overclock Video Cards, they are overclocked by manufactures by default, their default speed won't be (Stock) but mark as (OC Edition) in this benchmark. In order to make it (Stock) as original Video Card performance, you must downclock it to genuine Stock speed on its specification with Overclock Tools such as Evga Precision or MSI Afterburner.

    Unless the result at Stock speed of same model's Video Card as yours had existed already, you should at least post your benchmark result when your Video Card is not overclocked i.e.(Stock). Because there are people on Earth are unwilling to overclock, your results of not overclocked is helping them decide which Video Card would they purchase?
     
  3. Your user name at this forum
     
  4. GPU-Z (Optional): Link to GPU-Z screenshot of your Video Card
    I bet GPU-Z screenshot is helpful for those who are buying a Video Card. e.g. GPU-Z





DATABASE


NTSC PS2 BIOS - Mild Group (1024x1024)
Quote:231.88 FPS - NVIDIA GTX 460 SE (1 GB, 256 bit) (OC 850/925) - Rezard
225.35 FPS - AMD Radeon HD 6870 BE (OC 940/1150) - Rezard - GPU-Z
202.53 FPS - AMD Radeon HD 5850 (Stock 725/1000) - NRX - GPU-Z
192.77 FPS - NVIDIA GTX 460 SE (1 GB, 256 bit) (Stock 648/850) - Rezard - GPU-Z
164.10 FPS - NVIDIA GTX 550 Ti (OC 950/1075) - Apathy compleX - GPU-Z
156.10 FPS - NVIDIA GTX 550 Ti (Stock 900/1026) - Apathy compleX - GPU-Z
149.53 FPS - AMD Radeon HD 4870 (Stock 750/900, 512MB) - IcemanSR - GPU-Z
144.14 FPS - NVIDIA 9600GT (OC 700/1080, Driver 258.96 beta) - Shadow Lady - GPU-Z
134.45 FPS - NVIDIA 9600GT (OC 700/1080) - Shadow Lady - GPU-Z
132.78 FPS - AMD Radeon HD 5750 (OC 850/1200) - sakraycore - GPU-Z
122.61 FPS - AMD Radeon HD 5750 (Sapphire OC version 710/1160, Only OC 10 MHz each) - sakraycore
122.14 FPS - NVIDIA 9600GT (OC 730/1115) - CKL (invalid from post #129 see post #427 for details)
121.21 FPS - NVIDIA 9800GT (Stock 600/900, 512 MB) - retsnom - GPU-Z
119.85 FPS - NVIDIA 9600GT - Shadow Lady
116.36 FPS - NVIDIA GTX 560M (Stock 775/1250) - synce
114.70 FPS - NVIDIA 9600GT (OC 730/1000) - CKL
107.90 FPS (Aero theme, 82.5 FPS on basic theme) - NVIDIA GT 650M (2G GDDR3, stock 835/900) - avih - GPU-Z
106.67 FPS - AMD Radeon HD 5870M - yurihyuga
105.26 FPS - AMD Radeon HD 4850 (An OC Edition 650/1000, 512MB) - antespo - GPU-Z
105.26 FPS - NVIDIA 9600GT (BFG OC version 675/900) - CKL
100.00 FPS - AMD Radeon HD 4850 (Stock 625/993, 1GB) - iakoboss7 - GPU-Z
100.00 FPS - AMD Radeon HD 6670 (Stock 810/1000, GDDR5) - Rezard
93.84 FPS - AMD Radeon HD 5670 - ilovejedd - GPU-Z
82.90 FPS - NVIDIA GT 240 - ilovejedd - GPU-Z
79.40 FPS - AMD Radeon HD 3850 (OC 710/909) - Pedrosilvaa
73.56 FPS - AMD Radeon HD 3850 - Pedrosilvaa
66.81 FPS - NVIDIA GT 230M (OC 650/1040) - Rezard - GPU-Z
60.61 FPS - NVIDIA 8600GT (OC 756/900) - pcsx2fan - GPU-Z
60.26 FPS - NVIDIA GT 525M (OC 700/1000) - synce
57.76 FPS - NVIDIA GT 220 - Rezard - GPU-Z
55.75 FPS - NVIDIA GT 430 (Stock 700/800) - Rezard
51.61 FPS - NVIDIA GT 525M (Stock 600/900) - synce
51.53 FPS - NVIDIA GT 230M - Rezard - GPU-Z
50.00 FPS - AMD Radeon HD 6530D (Stock 444/933, DDR3 1866) - Rezard
48.71 FPS - Intel HD 4000 (Core i5 3570K) - Shadow Lady - GPU-Z
45.26 FPS - NVIDIA 8600GT - pcsx2fan - GPU-Z
42.40 FPS (basic theme, 41 on aero theme) - Intel HD Graphics 4000 (i7-3630QM, stock 650/667) - avih - GPU-Z
37.04 FPS - AMD Radeon HD 4670 (DDR2) - Secret Dragoon - GPU-Z
33.61 FPS - Intel HD 3000 (Core i7 2630QM) - thatwasonce
27.42 FPS - NVIDIA 9400GT - Rezard - GPU-Z
23.17 FPS - AMD Mobility Radeon HD 545v (Stock 680/747, DDR3) - Livy
19.54 FPS - Intel HD Graphics (Core i5 430M, DDR3) - Livy
17.07 FPS - AMD Radeon HD 6310 - Rezard
15.12 FPS - AMD Radeon HD 4350 (512MB) - ikazu
12.31 FPS - AMD Mobility Radeon HD 4200 - Rezard - GPU-Z
11.71 FPS - Intel G965 (Stock 667/0) - synce
9.56 FPS  - AMD Mobility Radeon HD 4200 (Stock 498/397, DDR2) - Livy
9.50 FPS  - NVIDIA 8200M G - Rezard - GPU-Z



NTSC PS2 BIOS - Harsh Group (2048x2048)
Quote:248.06 FPS - NVIDIA GTX 970 - Unr3al
201.26 FPS - NVIDIA GTX 780 Ti (OC 1006/1750, 3GB) - tuanming
174.86 FPS - AMD Radeon R9 290X (Asus OC Edition 1050/1350, 4GB) - tuanming
160.80 FPS - NVIDIA GTX 680 (OC 1312/1749) - tuanming - GPU-Z
160.80 FPS - NVIDIA GTX 670 (OC 1389/1907 where GPU Core in Turbo Boost) - hallmark - GPU-Z
153.11 FPS - NVIDIA GTX 770 (1137/1189, 4GB) - tuanming
145.45 FPS - NVIDIA GTX 670 (OC 1250/1750) - Unr3al - GPU-Z
140.35 FPS - NVIDIA GTX 760 (OC 1320/1822) - masterotaku
137.93 FPS - NVIDIA GTX 670 (EVGA FTW OC Edition 1229/1552 where GPU Core in Turbo Boost) - hallmark - GPU-Z
137.93 FPS - NVIDIA GTX 580 (OC 1000/1125) - tuanming
135.02 FPS - NVIDIA GTX 580 (OC 950/1100) - tuanming
134.45 FPS - NVIDIA GTX 680 (EVGA OC Edition 1020/1500, Only OC 14 MHz in Core) - tuanming - GPU-Z
130.08 FPS - NVIDIA GTX 670 - Shadow Lady - GPU-Z
125.49 FPS - NVIDIA GTX660 Ti (OC 1254/1824, ASUS DirectCU2) - Burr - GPU-Z
122.61 FPS - NVIDIA GTX 580 (MSI Twin Frozr III Lightning Extreme, OC Edition 832/1050) - tuanming
122.14 FPS - NVIDIA GTX 580 (Zotac OC Edition 815/1025) - tuanming
117.22 FPS - NVIDIA GTX 580 (Stock 772/1002, Driver 267.31 beta) - fbr1010 - GPU-Z
116.79 FPS - NVIDIA GTX 670 (OC 980/1502 where GPU Core in Turbo Boost) - hallmark - GPU-Z
112.28 FPS - NVIDIA GTX 480 (Gigabyte SOC Edition 821/950, Driver 267.24 beta) - fbr1010 - GPU-Z
102.24 FPS - AMD Radeon HD 6970 (OC 1000/1450) (Driver 11.8) - tuanming (invalid from post #417 see post #427 for details)
105.26 FPS - NVIDIA GTX 570 (1.2 GB, 320 bit) (OC 880/1010) - Flokk - GPU-Z
101.59 FPS - NVIDIA GTX 480 (Stock 700/924) - frosty5689 - GPU-Z
100.63 FPS - AMD Radeon HD 7850 (OC 1265/1400) (Driver 11.12 Beta8) - ||dav1de||
99.38 FPS - NVIDIA GTX 570 (1.2 GB, 320 bit) (EVGA Superclocked Edition 797/975) - Flokk - GPU-Z
99.07 FPS - NVIDIA GTX 480 (OC 700/950, SLI) - Unr3al - GPU-Z
96.68 FPS - NVIDIA GTX 470 (OC 800/1000) - yurihyuga - GPU-Z
96.39 FPS - NVIDIA GTX 285 (OC 712/1440, 1GB, Driver 266.58) - fbr1010 - GPU-Z
96.39 FPS - NVIDIA GTX 660 (Gigabyte Windforce, 2GB GDDR5, OC 1136/1502) - Bositman - GPU-Z
94.67 FPS - NVIDIA GTX 570 (1.2 GB, 320 bit) (Stock 732/950) - Flokk - GPU-Z
92.71 FPS - NVIDIA GTX 650 Ti (OC 1293/1700) - zeikai
91.69 FPS - NVIDIA GTX 560 Ti (OC 1000/1200) - Shadow Lady - GPU-Z
91.17 FPS - AMD Radeon HD 7870 GHz Edition (Gigabyte OC Edition 1100/1200) - Teentitanbg - GPU-Z
91.17 FPS - AMD Radeon HD 6950 (OC 1000/1425) - tuanming
90.91 FPS - AMD Radeon HD 6970 (Stock 880/1375) (Driver 11.4) - rebyc - GPU-Z
89.14 FPS - AMD Radeon HD 6950 (OC 987/1375) (Driver 11.1) - tuanming
88.76 FPS - NVIDIA GTX 650 Ti (OC 1293/1500) - zeikai - GPU-Z
88.40 FPS - NVIDIA GTX 560 Ti (OC 950/1200) - Shadow Lady - GPU-Z
86.96 FPS - NVIDIA GTX 470 (OC 730/893) - hallmark
86.25 FPS - NVIDIA GTX 285 (MSI OC Edition 680/1250, Driver 258.96) - fbr1010 - GPU-Z
85.11 FPS - AMD Radeon HD 6870 (OC 1050/1196) - cyber
82.90 FPS - AMD Radeon HD 5850 (OC 950/1200) - NRX - GPU-Z
82.05 FPS - AMD Radeon HD 5870 (OC 900/1300) - yurihyuga - GPU-Z
81.63 FPS - AMD Radeon HD 7850 (Stock 860/1200) (Driver 11.12 Beta8) - ||dav1de||
79.01 FPS - AMD Radeon HD 6950 (MSI OC Edition 810/1250, Only OC 10 MHz in Core) - tuanming
78.67 FPS - NVIDIA GTX 470 (An OC Edition 625/837, Only OC 18 MHz in Core) - yurihyuga - GPU-Z
78.43 FPS - NVIDIA GTX 470 (Stock 607/837) - hallmark
78.05 FPS - NVIDIA GTX 260 (User manually OC version 756/1269) - rama - GPU-Z
77.86 FPS - AMD Radeon HD 5870 (Stock 850/1200) - yurihyuga
76.74 FPS - AMD Radeon HD 6870 (ASUS 6870 DCU 915/1050, Only OC 15 MHz in Core) - cyber - GPU-Z
76.56 FPS - AMD Radeon HD 6870 BE (OC 940/1150) - Rezard - GPU-Z
75.43 FPS - NVIDIA GTX 560 Ti (Stock 822/1002) - Shadow Lady - GPU-Z
72.89 FPS - AMD Radeon HD 6850 (OC 900/1100) - synce
69.41 FPS - AMD Radeon HD 5850 (Stock 725/1000) - NRX - GPU-Z
66.81 FPS - AMD Radeon HD 5850 (GV-R585OC-1GD OC 765/1000, 1024MB) (Driver 10.10) - fbr1010 - GPU-Z
66.67 FPS - NVIDIA GTX 460 SE (1 GB, 256 bit) (OC 850/925) - Rezard
65.44 FPS - AMD Radeon HD 6850 (Stock 775/1000) - synce
65.44 FPS - NVIDIA GTX 460 (OC 825/1000) - thatwasonce - GPU-Z
64.78 FPS - AMD Radeon HD 5850 - xshadowinxbc
64.13 FPS - NVIDIA GTX 460 (1 GB, 256 bit) (OC 800/1000, Driver 258.96, GV-N460OC-1GI) - JTD
59.59 FPS - NVIDIA GTX 460 (768 MB, 192 bit) (OC 860/1010) - tuanming
58.72 FPS - NVIDIA GTX 460 (768 MB, 192 bit) (OC 832/1050) - cyber
55.56 FPS - NVIDIA GTX 460 (1 GB, 256 bit) (Stock 675/900, Driver 258.96, GV-N460OC-1GI) - JTD
55.27 FPS - NVIDIA GTX 460 SE (1 GB, 256 bit) (Stock 648/850) - Rezard - GPU-Z
52.55 FPS - NVIDIA GTX 460 (768 MB, 192 bit) (MSI OC Edition 751/900) - tuanming
51.04 FPS - AMD Radeon HD 5770 (OC 960/1425) - Air - GPU-Z (invalid from post #121 see post #427 for details)
49.31 FPS - AMD Radeon HD 4870 (Stock 750/900, 512MB) - IcemanSR - GPU-Z
45.71 FPS - NVIDIA GTS 250 (OC 738/1320) - ilovejedd
45.26 FPS - NVIDIA GTX 550 Ti (OC 950/1075) - Apathy compleX - GPU-Z
42.55 FPS - AMD Radeon HD 5750 (OC 850/1200) - sakraycore - GPU-Z
42.50 FPS - NVIDIA GTX 550 Ti (Stock 900/1026) - Apathy compleX - GPU-Z
41.83 FPS - AMD Radeon HD 5770 (Sapphire Vapor-X OC Edition 860/1200) (Driver 10.8) - fbr1010 - GPU-Z
39.56 FPS - AMD Radeon HD 5750 (Sapphire OC version 710/1160, Only OC 10 MHz each) - sakraycore
38.93 FPS - NVIDIA GTS 250 - ilovejedd
38.32 FPS - NVIDIA 9600GT (OC 700/1080, Driver 257.15 beta) - Shadow Lady - GPU-Z
36.41 FPS - NVIDIA 9600GT (OC 700/1080) - Shadow Lady - GPU-Z
35.71 FPS - NVIDIA 9600GT (OC 730/1115) - CKL (invalid from post #129 see post #427 for details)
34.86 FPS - AMD Radeon HD 5870M - yurihyuga
34.82 FPS - NVIDIA GTS 450 - ilovejedd
33.23 FPS - AMD Radeon HD 4850 (An OC Edition 650/1000, 512MB) - antespo - GPU-Z
32.65 FPS - NVIDIA 9800GT (Stock 600/900, 512 MB) - retsnom - GPU-Z
32.19 FPS - NVIDIA GTX 560M (Stock 775/1250) - synce
31.80 FPS - NVIDIA 9600GT (OC 730/1000) - CKL
31.65 FPS - NVIDIA 9600GT - Shadow Lady
31.25 FPS - AMD Radeon HD 4850 (Stock 625/993, 1GB) - iakoboss7 - GPU-Z
30.40 FPS (Aero theme. 27 FPS on basic theme) - NVIDIA GT 650M (2G GDDR3, stock 835/900) - avih - GPU-Z
29.93 FPS - NVIDIA 9600GT (BFG OC version 675/900) - CKL
29.82 FPS - AMD Radeon HD 5670 - ilovejedd - GPU-Z
24.10 FPS - AMD Radeon HD 3850 (OC 710/909) - Pedrosilvaa
23.63 FPS - NVIDIA GT 240 - ilovejedd - GPU-Z
22.33 FPS - AMD Radeon HD 3850 - Pedrosilvaa
18.22 FPS - NVIDIA GT 230M (OC 650/1040) - Rezard - GPU-Z
15.83 FPS - NVIDIA 8600GT (OC 756/900) - pcsx2fan - GPU-Z
14.88 FPS - NVIDIA GT 430 (Stock 700/800) - Rezard
13.65 FPS - NVIDIA GT 230M - Rezard - GPU-Z
12.46 FPS - Intel HD 4000 (Core i5 3570K) - Shadow Lady - GPU-Z
12.00 FPS - NVIDIA 8600GT - pcsx2fan - GPU-Z
10.22 FPS - AMD Radeon HD 4670 (DDR2) - Secret Dragoon - GPU-Z
10.20 FPS (basic theme, 8.7 FPS on aero theme) - Intel HD Graphics 4000 (i7-3630QM, stock 650/667) - avih - GPU-Z
8.63 FPS - Intel HD 3000 (Core i7 2630QM) - thatwasonce
6.37 FPS - AMD Mobility Radeon HD 545v (Stock 680/747, DDR3) - Livy
5.41 FPS - Intel HD Graphics (Core i5 430M, DDR3) - Livy
4.02 FPS - AMD Radeon HD 4350 (512MB) - ikazu
2.72 FPS - AMD Mobility Radeon HD 4200 (Stock 498/397, DDR2) - Livy



PAL PS2 BIOS - Mild Group (1024x1024)
Quote:104.92 FPS - AMD Radeon HD 4850 (Stock 625/993, 1GB) - iakoboss7 - GPU-Z
102.24 FPS - AMD Radeon HD 5670 (1GB, GDDR5) (GV-R567ZL-1GI OC 790/1000) - miseru99 - GPU-Z



PAL PS2 BIOS - Harsh Group (2048x2048)
Quote:121.67 FPS - NVIDIA GTX 480 (OC 850/1000, SLI) - ChronicNL - GPU-Z
117.22 FPS - NVIDIA GTX 480 (OC 800/1000, SLI) - ChronicNL - GPU-Z
105.26 FPS - NVIDIA GTX 570 (OC 900/1030, GDDR5) - refraction - GPU-Z
103.56 FPS - NVIDIA GTX 480 (OC 700/950, SLI) - Unr3al - GPU-Z
110.34 FPS - NVIDIA GTX 480 (OC 725/950, SLI) - ChronicNL - GPU-Z (invalid from post #237 see post #427 for details)
107.02 FPS - NVIDIA GTX 480 (SLI) - ChronicNL - GPU-Z (invalid from post #232 see post #427 for details)
98.46 FPS - NVIDIA GTX 660 (Gigabyte Windforce, 2GB GDDR5, OC 1136/1502) - Bositman - GPU-Z
91.69 FPS - NVIDIA GTX 570 (Stock 732/950, GDDR5) - refraction - GPU-Z
47.06 FPS - AMD Radeon HD 5750 (OC 870/1250) - sakraycore - GPU-Z
33.51 FPS - AMD Radeon HD 4850 (Stock 625/993, 1GB) - iakoboss7 - GPU-Z
30.42 FPS - AMD Radeon HD 5670 (1GB, GDDR5) (GV-R567ZL-1GI OC 790/1000) - miseru99 - GPU-Z



DQ8 (SLUS 21207) - Mild Group (1024x1024)
Quote:47.69 FPS - NVIDIA GT 230M (OC 650/1040) - Rezard - GPU-Z
47.48 FPS - NVIDIA 8600GT (OC 756/900) - pcsx2fan - GPU-Z
36.57 FPS - NVIDIA 8600GT - pcsx2fan - GPU-Z
35.71 FPS - NVIDIA GT 230M - Rezard - GPU-Z



DQ8 (SLUS 21207) - Harsh Group (2048x2048)
Quote:101.91 FPS (71% CPU shown in pcsx2 game window title) - NVIDIA GTX 670 (OC 1215/1750) - Unr3al
87.67 FPS (71% CPU shown in pcsx2 game window title) - NVIDIA GTX 580 (OC 950/1125) - tuanming

The results above are invalid and overkill, too powerful GPU so that CPU becomes bottleneck,
cuz FPS is limited by either CPU or GPU depending on which one is the slowest.


137.34 FPS - NVIDIA GTX 780 Ti (OC 1006/1750, 3GB) - tuanming
122.61 FPS - NVIDIA GTX 670 (OC 1306/1824) - Unr3al - GPU-Z
120.75 FPS - AMD Radeon R9 290X (Asus OC Edition 1050/1350, 4GB) - tuanming
109.59 FPS - NVIDIA GTX 770 (1137/1189, 4GB) - tuanming
95.24 FPS - NVIDIA GTX 580 (3GB, Lightning Extreme) - tuanming
90.40 FPS - NVIDIA GTX 670 (Stock 915/1500) - Unr3al - GPU-Z
82.69 FPS - NVIDIA GTX 570 (OC 900/1030, GDDR5) - refraction - GPU-Z
82.05 FPS - NVIDIA GTX 480 (OC 700/950, SLI) - Unr3al - GPU-Z
70.95 FPS - NVIDIA GTX 570 (Stock 732/950, GDDR5) - refraction - GPU-Z
65.17 FPS - AMD Radeon HD 6970 (OC 1000/1450) (Driver 11.8) - tuanming
57.66 FPS - AMD Radeon HD 6950 (OC 1000/1425) - tuanming
57.55 FPS - NVIDIA GTX 260 (User manually OC version 756/1269) - rama - GPU-Z
57.45 FPS - AMD Radeon HD 6950 (OC 987/1375) (Driver 11.1) - tuanming
50.79 FPS - AMD Radeon HD 6950 (MSI OC Edition 810/1250, Only OC 10 MHz in Core) (Driver 11.1) - tuanming
47.34 FPS - NVIDIA GTX 460 (768 MB, 192 bit) (OC 860/2020) - tuanming
32.16 FPS - AMD Radeon HD 5750 (OC 850/1200) - sakraycore - GPU-Z
21.11 FPS - NVIDIA 9800GT (Stock 600/900, 512 MB) - retsnom - GPU-Z
20.43 FPS - NVIDIA 9600GT (OC 730/1115)- CKL
16.33 FPS - NVIDIA 9600GT (BFG OC version 675/900) - CKL
5.52 FPS - NVIDIA 8600GT - pcsx2fan - GPU-Z



Appendix
  1. different os won't affect pcsx2 game speed: I tested no speed changes among Windows 7(x86 and x64) and Vista(x86).
  2. All Video Cards' specifications from Wikipedia
  3. Credits: pcsx2fan (author)
durable PC since 2008
CPU : E7200 @ 2.53 GHz 1.04 V / OC 3.7 GHz 1.36 V (390*9.5, 4GB DDR2 780)
GPU : 8600GT (GDDR3, 256 MB) / OC (750 / 900)
BOARD : GA-EP31-DS3L (rev 1.0) broken and bought a second hand one replaced
PSU : 300 W
OS : Windows 7 Home Basic x86
Reply

Sponsored links

#2
Nice setup, pcsx2fan! Can't wait to contribute. Laugh

Just thought I'd mention, the 3 more important parts of a GPU as far as PCSX2 are in order from most to least important.

1. Bus Width
2. Fill Rate
3. Video RAM

I believe that's correct. I think I saw a post by Air stating this. I'll see if I can find that thread/post to double check, if anyone feels I'm off my rocker.

BTW: I'm only mentioning these as information in a relavant place. A little FYI for those who don't know, Smile
Reply
#3
In this case I think just using a GSdump and running it with the GSDumpGui (from the PCSX2 SVN) would be much better and easier.

I'm also not sure if you're supposed to distribute save states either Tongue
Core i5 3570k -- Geforce GTX 670  --  Windows 7 x64
Reply
#4
yep pal bios run's at 50 fps and ntsc at 60fps so probably u have to create 2 categories like u did with pal ffx-2 and ntsc ffx-2
OS: Win 7 Ultimate x64 sp1, MoBo: Asus P5QD Turbo, CPU: Q6600 @ 3,0Ghz, RAM: Trancend 2x2gb 6-6-6-18 800 MHz, GPU: HD 4850 1gb.
Pcsx2: Always Latest
Reply
#5
(05-16-2010, 09:09 PM)Shadow Lady Wrote: I'm also not sure if you're supposed to distribute save states either Tongue

Why ya gotta break balls? Tongue

Let me guess, "it's the rules". LOL
Reply
#6
(05-16-2010, 08:16 PM)Rezard Wrote: Just thought I'd mention, the 3 more important parts of a GPU as far as PCSX2 are in order from most to least important.

1. Bus Width
2. Fill Rate
3. Video RAM

I believe that's correct. I think I saw a post by Air stating this. I'll see if I can find that thread/post to double check, if anyone feels I'm off my rocker.

Everyone had a measurement or judgement bearing in their mind towards Video Card. So it's unnecessary to supply GPU's specifications just the Name of Video Card is enough.

I'll change this line
Quote:I've no idea what of the GPU's specifications ought to be entered. I suggest you should post GPU-Z screenshot instead.

to this one
Quote:you should post GPU-Z screenshot, I bet this information is helpful for those who are buying a Video Card.
durable PC since 2008
CPU : E7200 @ 2.53 GHz 1.04 V / OC 3.7 GHz 1.36 V (390*9.5, 4GB DDR2 780)
GPU : 8600GT (GDDR3, 256 MB) / OC (750 / 900)
BOARD : GA-EP31-DS3L (rev 1.0) broken and bought a second hand one replaced
PSU : 300 W
OS : Windows 7 Home Basic x86
Reply
#7
pal bios 1024x1024: 104.92 FPS - ATI HD 4850 1GB - iakoboss7 - GPU-Z

pal bios 2048x2048: 33.51 FPS - ATI HD 4850 1GB - iakoboss7 - GPU-Z

ntsc bios 1024x1024: 100 FPS - ATI HD 4850 1GB - iakoboss7 - GPU-Z

ntsc bios 2048x2048: 31.25 FPS - ATI HD 4850 1GB - iakoboss7 - GPU-Z

i didnt used a lower resolution in the monitor cause it will ***** up things in desktop etc. also i used the resolution u said in pcsx2 window (at 60 hz option cause i got a 75 one also) and kept it as it is cause if i maximise it to my screen's full res then the fps drop and that wouldnt be accurate cause everyone have different resolution in their monitor! if u think i did it wrong u can ask me to run them again.!


Attached Files Thumbnail(s)
               
OS: Win 7 Ultimate x64 sp1, MoBo: Asus P5QD Turbo, CPU: Q6600 @ 3,0Ghz, RAM: Trancend 2x2gb 6-6-6-18 800 MHz, GPU: HD 4850 1gb.
Pcsx2: Always Latest
Reply
#8
I also recommend you scratch the game tests, pcsx2fan.
It's far too complex to do, and the requirement of uploading lots of savestates isn't nice either.
The BIOS however is more reasonable, since everyone has it and it's easy to reach.

Edit:
Doing these tests and building a database on it isn't nearly as helpful as the CPU one though.
Actual games have different and constantly changing requirements.
The BIOS is a good example of this, reaching 400fps normally, but slowing down to a mere 50fps when the PS2 logo fades in.
Even worse is when your card runs out of memory, which happens fast with GSdx in some games / internal rendering resolution..
Reply
#9
(05-16-2010, 10:04 PM)rama Wrote: I also recommend you scratch the game tests, pcsx2fan.
It's far too complex to do, and the requirement of uploading lots of savestates isn't nice either.
The BIOS however makes for a good test, since everyone has it and it's easy to reach.

yep i think just the bios itself is enough and that way everyone will be able to post his/her results and will not be limited to ppl that own a game and will be a little more simple to do probably. having 4 categories already cause of ntsc-pal bios is enough Tongue

edit: about for when the ps2 logo fades in isnt this because of the effect it creates? like i have seen the same thing for fog, dust, light effects. why is it happening? (u can edit your post so we dont get offtopic if that will be a problem)

edit2: the screen size option in the bios menu doesnt affect the speed at all so i dont think anyone should change it (i mentioned it cause i saw u used the full option and i use the 4:3)
OS: Win 7 Ultimate x64 sp1, MoBo: Asus P5QD Turbo, CPU: Q6600 @ 3,0Ghz, RAM: Trancend 2x2gb 6-6-6-18 800 MHz, GPU: HD 4850 1gb.
Pcsx2: Always Latest
Reply
#10
Mild
114.70 FPS - NVIDIA 9600GT OC 730mhz memory 1000mhz- CKL
102.56 FPS - NVIDIA 9600GT (BFG oc version stock clock 675mhz memory 900mhz) - CKL
Harsh
31.80 FPS - NVIDIA 9600GT OC 730mhz memory 1000mhz - CKL
28.5 FPS - NVIDIA 9600GT (BFG oc version stock clock 675mhz memory 900mhz) - CKL
NTSC bios
AMD Athlon II 245 @ 3.55ghz, 9600GT @730mhz 1115mhz, vista sp2 32bit
Reply




Users browsing this thread: 1 Guest(s)