iXBT Labs - Computer Hardware in Detail

Platform

Video

Multimedia

Mobile

Other

August 2002 3Digest



Monthly drivers report and popular 3D accelerators comparison

Monthly 3D Digest is the project of iXBT.com devoted to releases of new Windows XP (monthly) and Windows 9x/ME (once in 4 months) drivers for popular videocards (released from 10th of July 2002 to 10th August 2002). As usual, there are many diagrams containing the information on changes in operation of videocards at change to last version drivers, and also summary diagrams of popular modern videocards performance. As the tradition of our 3D Digest there is the best price/performance ratios information. Quality aspects are reviewed in the Windows XP section.

It's our 27th issue of 3Digest.

Contents

  1. Testbed configuration and test software list
  2. Videocard list
  3. What's new in August 2002 3Digest
  4. Test summary diagrams and digest bottom line
  5. Anisotropic filtering examination
  6. Minimal and maximum FPS research with Comanche4 (July 2002)
  7. Examination of performance difference between Intel Pentium4: i850 and i845D platforms (April 2002)
  8. 3D accelerator rating calculations for the end of August 2002
  9. Screenshot gallery







1. Testbed configuration



1.1. Pentium 4 2200 MHz based computer:
  • CPU: Intel Pentium 4 2200 (L2=512K);
  • Mainboard: ASUS P4T-E (i850);
  • RAM: 512 MB RDRAM PC800;
  • HDD: Quantum FB AS 20GB;



1.3. Athlon XP 2000+ based computer:
  • CPU: AMD Athlon XP 2000+ (1666 MHz);
  • Mainboard: EpoX 8KHA+ (VIA KT266A);
  • RAM: 512 MB DDR SDRAM PC2100;
  • HDD: Seagate Barracuda IV 40GB;





The following additional testbed was used for the platform performance comparison:

1.5. Pentium 4 2200 MHz based computer:
  • CPU: Intel Pentium 4 2200;
  • Mainboard: Leadtek P4845D (i845D);
  • RAM: 512 MB DDR SDRAM PC2100;
  • HDD: Quantum FB AS 20GB;





All computers had Windows ME and Windows XP, DirectX 9.0 beta 2 installed. ViewSonic P810 (21") and ViewSonic P817-E (21") monitors were used.

These games were used as testing instruments:

  • Return to Castle Wolfenstein (MultiPlayer) (id Software/Activision) - OpenGL, multitexturing, Checkpoint-demo, best test settings, S3TC OFF, configs are here

    Testing order: invoke console by pressing '~' and type

    timedemo 1 (press Enter)
    demo checkpoint (press Enter)


  • 3DMark2001 SE Pro (MadOnion/Remedy), Game2 "Dragothic" - DirectX 8.0, Hardware TCL, multitexturing, LOW Details, DXTC OFF, double buffering 24-bit Z-buffer

    Testing order:

    Through 3DMark2001 settings menu
    in Batch mode


  • Serious Sam: The Second Encounter v.1.05 (Croteam/GodGames) - OpenGL, multitexturing, Grand Cathedral demo, test settings: quality, S3TC OFF

    Testing order: invoke console by pressing '~' and type

    dem_bProfile=1 (press Enter)
    Run Grand Cathedral demo from the menu


  • 3DMark2001 Pro (MadOnion/Remedy) - DirectX 8.0, Hardware TCL, Game1, Game4 Game4 Game4, Low, High detail levels

  • Comanche4 Benchmark Demo (NovaLogic) - Direct3D, Shaders, Hardware T&L, Dot3, cube texturing, maximum quality

    Testing order: run the game this way

    c4demo.exe /advbench fps.txt 1024 32 noaudio nct dx7 (press Enter)


  • Unreal Tournament 2003 Demo v.927 (Digital Extreme/Epic Games) - Direct3D, Vertex Shaders, Hardware T&L, Dot3, cube texturing, default quality

    Testing order: run benchmark.bat from the System game folder, changing resolution only.

  • RightMark Video Analyzer v.0.4 (Philip Gerasimov) - DirectX 8.1, Dot3, cube texturing, shadow buffers, vertex and pixel shaders (1.1, 1.4).

    Test settings: pixel shaders 1.1, shadow buffers OFF.

2. Videocard list

Here is the list of graphic processors (videocards) tested, clicking on every name you'd be taken to a page with a report about new drivers for the card based on the appropriate chip and a short list of card's features. Summary diagrams also contain results for overclocked videocards.




Testing order:

  1. New drivers are tested on the Pentium4 2200 MHz testbed;
  2. Summary diagrams of comparative videocards' performance are formed using all the testbeds.

Drivers released by videocard manufacturers are not taken into account (unless written by GPU manufacturers) because there are very many of them and it's impossible to test them all.




You can find comments about reviewed drivers in the corresponding sections (see the list below). 3D graphics quality was tested on the following software:

  • Return to Castle Wolfenstein (id Software/Activision)
  • Unreal 2 (Epic MegaGames/Legend Entertainment)
  • Comanche4 (Novalogic)
  • Unreal Tournament 2003 (Digital Extreme/Epic Games)
  • Colin McRae Rally 2 (Codemasters)
  • Red Faction (Volition/THQ)
  • realMyst (Cyan Worlds/Mattel Interactive)
  • Rally Trophy (Bugbear Entertainment/JoWood Productions)
  • Sacrifice (Shiny/Interplay)
  • Codename: Outbreak (Venom)(GSC Game World/Russobit-M/Virgin Interactive Entertainment)
  • Severance: Blade of Darkness (Rebel Act Studios/Codemasters)
  • Neverwinter Nights (Infogrames/Bioware)
  • Warcraft III (Blizzard)
  • Soldier of Fortune II: Double Helix (Activision/Raven Software)
  • IL2: Shturmovik (1C Maddox Games/UbiSoft)
  • Max Payne (Remedy/GodGames)
  • 3D Mark 2001 Pro (Madonion.com/Remedy)
  • AquaNox (Fishtank/Massive Development)
  • Serious Sam: The Second Encounter (Croteam/GodGames)
  • Command & Conquer: Renegade (Westwood Studios/EA Games)
  • Medal of Honor: Allied Assault (EA Games)
  • Elder Scrolls III: Morrowind (Bethesda Softworks)
  • Star Wars: Jedi Knight II - Jedi Outcast (Activision/Lucas Arts)

Don't be offended if your favourite game is not in the list. We'll expand the list of games as fast as possible.

Card list:

  1. NVIDIA GeForce2 MX400 64MB SDR 6ns (200/183 MHz)
  2. NVIDIA GeForce2 Ti 64MB DDR 5ns (250/400 MHz)
  3. NVIDIA GeForce3 Ti VX 64MB DDR 5ns (225/366 MHz) - will be placed in the archive in August 2002
  4. NVIDIA GeForce4 MX 420 64MB SDR 6ns (250/166 MHz)
  5. NVIDIA GeForce4 MX 440 64MB DDR 5ns (270/400 MHz)
  6. NVIDIA GeForce4 MX 440 Ultra 64MB DDR 3.6ns (270/500 MHz)
  7. NVIDIA GeForce4 MX 460 64MB DDR 3.6ns (300/550 MHz)
  8. NVIDIA GeForce3 Ti 500 64MB DDR 3.8ns (240/500 MHz)
  9. NVIDIA GeForce3 Ti 200 64MB DDR 5ns (175/400 MHz)
  10. NVIDIA GeForce4 Ti 4200 64MB DDR 4ns (250/513 MHz)
  11. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/444 MHz)
  12. NVIDIA GeForce4 Ti4200 128MB DDR 3.3ns (250/550 MHz)
  13. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/500 MHz)
  14. NVIDIA GeForce4 Ti 4400 128MB DDR 3.6ns (275/550 MHz)
  15. NVIDIA GeForce4 Ti 4600 128MB DDR 2.8ns (300/650 MHz)
  16. NVIDIA GeForce4 Ti 4600 GS 128MB DDR 2.8ns (310/680 MHz)
  17. ATI RADEON 7500 64MB DDR 5.5ns (230/333 MHz)
  18. ATI RADEON 7500 64MB DDR 4ns (265/390 MHz)
  19. ATI RADEON 7500 64MB DDR 4ns (270/460 MHz)
  20. ATI RADEON 7500 64MB DDR 4ns (290/460 MHz)
  21. ATI RADEON 8500 128MB DDR 3.3ns (300/600 MHz)
  22. ATI RADEON 8500 128MB DDR 3.6ns HYNIX (275/550 MHz)
  23. ATI RADEON 8500 128MB DDR 3.6ns SAMSUNG (275/550 MHz)
  24. ATI RADEON 8500 64MB DDR 3.6ns (275/550 MHz)
  25. ATI RADEON 8500LE 64MB DDR 5ns (183/366 MHz)
  26. ATI RADEON 8500LE 64MB DDR 4ns (230/460 MHz)
  27. ATI RADEON 8500LE 64MB DDR 5ns (250/366 MHz)
  28. ATI RADEON 8500LE 64MB DDR 3.6ns (250/500 MHz)
  29. ATI RADEON 8500LE 128MB DDR 4ns (250/500 MHz)
  30. ATI RADEON 9000 128MB DDR 3.6ns (250/400 MHz)
  31. ATI RADEON 9000 Pro 128MB DDR 3.6ns (275/550 MHz)
  32. ATI RADEON 9000 Pro 64MB DDR 3.6ns (275/550 MHz)
  33. ATI RADEON 9700 Pro 128MB DDR 2.8ns (325/620 MHz)
  34. PowerVR/STM KYRO II 64MB SDR 5ns (175/175 MHz)
  35. PowerVR/STM KYRO II SE 64MB SDR 4.7ns (200/200 MHz)
  36. SIS Xabre 400 64MB DDR 3.5ns (250/500 MHz)
  37. SIS Xabre 200 64MB DDR 4ns (200/366 MHz)
  38. SIS Xabre 400 128MB DDR 4ns (250/500 MHz)
  39. SIS Xabre 600 64MB DDR 4ns (275/550 MHz)
  40. SIS Xabre 600 128MB DDR 4ns (275/550 MHz)
  41. Matrox Parhelia 128MB DDR 3.3ns (220/550 MHz) (Retail)
  42. 3DLabs VPU (P10), Wildcat VP870 128MB DDR 3.3ns (~230/550 MHz)





Cards coming in the next 3Digest:

  1. S3 SavageXP 64MB DDR 5ns (166/366 MHz);
  2. Matrox Parhelia 128MB DDR 3.3ns (200/500 MHz) (OEM).
  3. Trident Blade XP4 T2 64MB DDR.





Archive (videocards, which info is no longer updated)




3. August 2002 news and the current business

  • 3.1. The following drivers were released from July 10, 2002 to August 10, 2002:

    • New Detonator XP (4) series from NVIDIA, for videocards based on Riva TNT, Riva TNT2M64/Vanta, TNT2/Pro/Ultra, GeForce256 SDR/DDR, GeForce2 GTS/MX/Pro/Ultra/Ti, GeForce3/Ti 200/500, GeForce 4 Ti/MX versions:



    • From ATI, for the videocard based on RADEON 9700, versions:

      • Windows XP: 6.143


    • From Matrox, for Parhelia-based cards, versions:

  • 3.2. We remind that from the 21st issue of 3Digest we test cards in 32-bit color mode only!

  • 3.3. From July 2002 Quake3 was replaced by Unreal Tournament 2003 DEMO build 927.

  • 3.4. From this issue all cards will be tested on 3 testbeds (without price separation). This will enable to demonstrate the performance of high-end cards on a relatively low-performance Pentium III-1000. We hope this will clearly explain the situation to users trying to install high-end solutions on low-end PCs. Due to the increased amount of work, we cancelled monthly Windows ME testing, making them once per 4 months (especially as all drivers optimizations are made for Windows XP first, and some companies (like Matrox) do not release Windows 9x/ME drivers at all. The next Windows ME testing will be conducted in October 2002.

  • 3.5. Anisotropic filtering research was considerably updated in July 2002. We added some screenshots taken on new cards.

  • 3.6. 3DMark2001 SE was updated in July 2002 and conducted on two platforms: Pentium4 and Athlon XP. Moreover, in this section we compared the videocards performance using hardware and software T&L (via CPU). The next 3DMark2001 SE testing will be conducted in September 2002.

  • 3.7. We examined the performance difference of a number of leading 3D-accelerators between Intel Pentium4: i850 and i845D platforms in April. The results are here. We have already examined the performance difference between Intel Pentium4 i850 and VIA P4X266.

  • 3.8. From this issue drivers are tested at 1024x768, 1280x1024, 1600x1200 resolutions.

  • 3.9. In July issue we used our Pentium4 testbed to conduct the examination of minimal and peak FPS using Comanche4 DEMO (Windows XP). You can take a look at the same results in the section 4, separated by resolutions.

    Note: at examination of minimal performance values one mustn't consider once-only FPS falloffs on all accelerators as it might be due to activation of some operating system functions that could use some CPU resources for a short time. Only oft-recurring minimal performance values can be used for gameplay estimation of different videocards at different resolutions.

  • 3.10. We added 3DLabs Wildcat VP870 to our tests, despite its professional intent. Creative, the new owner of 3DLabs, announced plans to release a gaming videocard based on VPU P10. So it would be very interesting to compare current test results of drivers, optimized for professional applications, with results of game-optimized P10 drivers.

  • 3.11. The test tools list was updated with the new RightMark Video Analyzer test, created by iXBT.com. We presented the first beta version in the NVIDIA GeForce4 Ti review. In this Digest we used 0.4 version using pixel (1.1 and 1.4 versions) and vertex shaders, and shadow buffers. We used 1.1 shaders and disabled shadow buffer to fully compare NVIDIA GeForce3,4 and ATi RADEON 9700/8500/LE.






4. Summary diagrams of videocards performance with the latest drivers for August 2002

Overclocked videocards are marked red. Clock rates follow the 'o/c' sign.



5. Anisotropic filtering investigation on the example of Serious Sam:TSE (OpenGL), RtCW (OpenGL), and 3DMark2001 SE (Direct3D) (Windows XP)

6. Examination of minimal and peak FPS with Comanche4 DEMO (Windows XP)

7. Examination of performance difference between Intel Pentium4: i850 and i845D platforms

8. 3D accelerator rating calculations for the end of August 2002



My thanks to:
Maxxx (maxm@online.sinor.ru),
Mikhail Sugakevitch (Mishail@newmail.ru)
Vyacheslav Gordeev AKA Slaydev (sarz@dale.elektra.ru) and
Ruslan73 (rusik73@yahoo.com)
dmitro13 (dmytro13@hotmail.com)
UnVial (lvg@pop.ioffe.rssi.ru)
for improving usability coefficient calculation method.

Anyone who wants to conduct his own rating calculations with his own values may download this Excel file (Office'97-'2000) in RAR 3.0 and ZIP files.

Calculations were conducted considering the following assumptions:

  1. Windows XP is used;
  2. Percentage of performance and quality demand ratio:
    • performance priority - 60%
    • quality priority - 40%
  3. ONLY 1600x1200@32bpp resolution was used for the prospect rating;
  4. Usability ratings are separated into the three sections: High-End, Middle-End, and Low-End. Each rating features its own expenses for the system upgrade. It's considered in APG parameter, that's 80 for Low-End, 180 for Middle-End, and 380 for High-End. We believe that if a user wants to purchase a very expensive and powerful accelerator, he understands the inanity of installing it into a Low-End PC. So we separated the usability rating for different PC levels.

* All speed values were obtained from tests, conducted on Pentium4 2200 MHz testbed.

8.1. 3D accelerator usability and prospect ratings for the end of August 2002

Usability ratings are calculated by the following formula:

Kusa = (((W10xp*0.6 + W12xp*0.3 + W16xp*0.1) +
      (UT10xp*0.6 + UT12xp*0.3 + UT16xp*0.1) +
      (C410xp*0.6 + C412xp*0.3 + C416xp*0.1) +
      (G210xp*0.6 + G212xp*0.3 + G216xp*0.1) +
      (SS10xp*0.6 + SS12xp*0.3 + SS16xp*0.1) +
      (3DMG1L10xp*0.6 + 3DMG1L12xp*0.3 + 3DMG1L16xp*0.1) +
      (3DMG1H10xp*0.6 + 3DMG1H12xp*0.3 + 3DMG1H16xp*0.1) +
      (3DMG2L10xp*0.6 + 3DMG2L12xp*0.3 + 3DMG2L16xp*0.1) +
      (3DMG2H10xp*0.6 + 3DMG2H12xp*0.3 + 3DMG2H16xp*0.1) +
      (3DMG3H10xp*0.6 + 3DMG3H12xp*0.3 + 3DMG3H16xp*0.1) +
      (3DMG410xp*0.6 + 3DMG412xp*0.3 + 3DMG416xp*0.1))* 3 * 0.6 +
      (Poss * 0.4)) *2 * 4/ (Price + APG)

The similar calculation (but for the 1600x1200 resolution only) indicates the prospect of a videocard. The formula contains the same indexes as those used for usability rating calculations, but accelerator's price isn't considered. So, this rating indicates the capabilities of current videocards in current applications and future games.

Kpro = ((W16xp +
      UT16xp +
      C416xp +
      G216xp +
      SS16xp +
      3DMG1L16xp +
      3DMG1H16xp +
      3DMG2L16xp +
      3DMG2H16xp +
      3DMG3H16xp +
      3DMG416xp) * 0.6 +
      (Poss * 0.4)) * 2 / 10

You can also compare two ratings to see the correspondence of product prices to their capabilities and features.

Where:

  • Kusa - usability rating. The higher it is, the better card's mark is.
  • Kpro - prospect rating. The higher it is, the better card's mark is.
  • W10xp - speed in Return to Castle Wolfenstein (checkpoint) at 1024X768X32 (Windows XP);
  • W12xp - speed in Return to Castle Wolfenstein (checkpoint) at 1280X1024X32 (Windows XP);
  • W16xp - speed in Return to Castle Wolfenstein (checkpoint) at 1600X1200X32 (Windows XP);
  • UT10xp - speed in UT2003 at 1024X768X32 (Windows XP) at maximum quality;
  • UT12xp - speed in UT2003 at 1280X1024X32 (Windows XP) at maximum quality;
  • UT16xp - speed in UT2003 at 1600X1200X32 (Windows XP) at maximum quality;
  • C410xp - speed in Comanche4 DEMO at 1024X768X32 (Windows XP);
  • C412xp - speed in Comanche4 DEMO at 1280X1024X32 (Windows XP);
  • C416xp - speed in Comanche4 DEMO at 1600X1200X32 (Windows XP);
  • G210xp - speed in 3DMark2001, Game2 Low at 1024X768X32 (Windows XP);
  • G212xp - speed in 3DMark2001, Game2 Low at 1280X1024X32 (Windows XP);
  • G216xp - speed in 3DMark2001, Game2 Low at 1600X1200X32 (Windows XP);
  • SS10xp - speed in Serious Sam at 1024X768X32 (Windows XP);
  • SS12xp - speed in Serious Sam at 1280X1024X32 (Windows XP);
  • SS16xp - speed in Serious Sam at 1600X1200X32 (Windows XP);
  • 3DMG1L10xp - speed in 3DMark2001 Game1 Low at 1024x768x32 (Windows XP);
  • 3DMG1H10xp - speed in 3DMark2001 Game1 High at 1024x768x32 (Windows XP);
  • 3DMG2L10xp - speed in 3DMark2001 Game2 Low at 1024x768x32 (Windows XP);
  • 3DMG2H10xp - speed in 3DMark2001 Game2 High at 1024x768x32 (Windows XP);
  • 3DMG3H10xp - speed in 3DMark2001 Game3 High at 1024x768x32 (Windows XP);
  • 3DMG410xp - speed in 3DMark2001 Game4 at 1024x768x32 (Windows XP);
  • 3DMG1L12xp - speed in 3DMark2001 Game1 Low at 1280x1024x32 (Windows XP);
  • 3DMG1H12xp - speed in 3DMark2001 Game1 High at 1280x1024x32 (Windows XP);
  • 3DMG2L12xp - speed in 3DMark2001 Game2 Low at 1280x1024x32 (Windows XP);
  • 3DMG2H12xp - speed in 3DMark2001 Game2 High at 1280x1024x32 (Windows XP);
  • 3DMG3H12xp - speed in 3DMark2001 Game3 High at 1280x1024x32 (Windows XP);
  • 3DMG412xp - speed in 3DMark2001 Game4 at 1280x1024x32 (Windows XP);
  • 3DMG1L16xp - speed in 3DMark2001 Game1 Low at 1600X1200X32 (Windows XP);
  • 3DMG1H16xp - speed in 3DMark2001 Game1 High at 1600X1200X32 (Windows XP);
  • 3DMG2L16xp - speed in 3DMark2001 Game2 Low at 1600X1200X32 (Windows XP);
  • 3DMG2H16xp - speed in 3DMark2001 Game2 High at 1600X1200X32 (Windows XP);
  • 3DMG3H16xp - speed in 3DMark2001 Game3 High at 1600X1200X32 (Windows XP);
  • 3DMG416xp - speed in 3DMark2001 Game4 at 1600X1200X32 (Windows XP);
  • Price - card price for the end of the month, according to price lists of some popular companies (for an average card);
  • Poss - estimated functional potential and quality of a videocard (see remarks below);

Poss is calculated by the following formula:

Poss = (2D*10 + Driver*70 + AF + TLF + DOT3 + EMBM + CEM + AA + S3TC + Glide + VS + PS + TCL + DH)/4

Where:

  • 2D - estimated 2D graphics quality on the 100-mark scale;
  • Driver - estimated drivers quality (including installation difficulties, etc.) on the 100-mark scale;
  • AF - anisotropic filtering:
    • = 700 - RADEON 9700;
    • = 600 - GeForce3/4;
    • = 500 - RADEON, RADEON 8500/9000;
    • = 400 - GeForce2, Parhelia;
    • = 150 - KYRO, KYRO II;
    • = 0 - n/a;
  • TLF - trilinear filtering:
    • = 600 - available;
    • = 200 - potentially possible, but with great resource losses;
    • = 150 - approximation;
  • DOT3 - Bump Mapping DOT3:
    • = 600 - available;
    • = 0 - n/a;
  • EMBM - Bump Mapping EMBM:
    • = 600 - available;
    • = 0 - n/a;
  • CEM - Cubic Environment Mapping:
    • = 600 - available;
    • = 0 - n/a;
  • AA - antialiasing (on condition of minimal gameplay):
    • = 700 - GeForce4 MSAA, Parhelia FAA16x, RADEON 9700 MSAA ;
    • = 600 - GeForce3 MSAA;
    • = 400 - SSAA, RADEON 8500 SmoothVision;
    • = 0 - n/a;
  • S3TC - texture compression support:
    • = 600 - KYRO, KYRO II, RADEON, RADEON 7500, RADEON 8500/9000/9700, Parhelia, GeForce4;
    • = 400 - GeForce256/2/3;
    • = 0 - n/a;
  • Glide - Glide support for UT-engine games:
    • = 450 - available;
    • = 0 - n/a;
  • VS - vertex shaders hardware support
    • = 700 - Parhelia, RADEON 9700;
    • = 400 - GeForce3/4, RADEON 8500;
    • = 0 - n/a;
  • PS - pixel shaders hardware support
    • = 800 - version 2.0;
    • = 600 - version 1.4;
    • = 450 - version 1.3;
    • = 400 - version 1.1;
    • = 0 - n/a;
  • TCL - estimated Hardware TCL:

    • = 400 - available;
    • = 0 - n/a;
  • DH - additional features: dual-monitor output with the support of DualHead/TwinView/HydraVision technologies, TV-tuner, etc.;

    • = 4000 - All-in-Wonder RADEON 8500DV/7500DV;
    • = 2100 - Matrox Parhelia;
    • = 700 - RADEON 7500, RADEON 8500/9000/9700, GeForce4;
    • = 500 - Matrox G400/G450/G550;
    • = 400 - GeForce2 MX (not taken into account, as cards supporting TwinView are more expensive);
    • = 0 - n/a;


For a table, download this Excel file (Office'97-'2000).

"Usability rating" (rating from the previous 3Digest is in parentheses):

High-End usability rating

  1. NVIDIA GeForce4 Ti 4200 64MB DDR 4ns (250/513 MHz) (1)
  2. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/500 MHz) (10)
  3. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/444 MHz) (2)
  4. NVIDIA GeForce3 Ti 500 64MB DDR 3.8ns (240/500 MHz) (7)
  5. ATI RADEON 8500LE 64MB DDR 3.6ns (250/500 MHz) (8)
  6. NVIDIA GeForce4 Ti 4400 128MB DDR 3.6ns (275/550 MHz) (3)
  7. ATI RADEON 8500 64MB DDR 3.6ns (275/550 MHz) (15)
  8. NVIDIA GeForce3 Ti 200 64MB DDR 5ns (175/400 MHz) (4)
  9. NVIDIA GeForce4 Ti 4200 128MB DDR 3.3ns (250/550 MHz) (6)
  10. ATI RADEON 8500LE 128MB DDR 3.6ns (250/500 MHz) (12)
  11. ATI RADEON 9000 Pro 64MB DDR 3.6ns (275/550 MHz) (5)
  12. NVIDIA GeForce4 Ti 4600 128MB DDR 2.8ns (300/650 MHz) (11)
  13. ATI RADEON 9700 128MB DDR 2.8ns (325/620 MHz) (-)
  14. ATI RADEON 8500LE 64MB DDR 4ns (230/460 MHz) (14)
  15. SIS Xabre600 64MB DDR 4ns (275/550 MHz) (17)
  16. ATI RADEON 8500 128MB DDR 3.6ns Hynix (275/550 MHz) (18)
  17. ATI RADEON 8500LE 64MB DDR 5ns (250/366 MHz) (16)
  18. ATI RADEON 9000 Pro 128MB DDR 3.6ns (275/550 MHz) (13)
  19. SIS Xabre600 128MB DDR 4ns (275/550 MHz) (19)
  20. SIS Xabre400 64MB DDR 3.5ns (250/500 MHz) (20)
  21. NVIDIA GeForce4 MX 460 64MB DDR 3.6ns (300/550 MHz) (23)
  22. SIS Xabre400 128MB DDR 4ns (250/500 MHz) (21)
  23. NVIDIA GeForce4 MX 440 64MB DDR 5ns (270/400 MHz) (24)
  24. ATI RADEON 8500LE 64MB DDR 5ns (183/366 MHz) (26)
  25. ATI RADEON 7500 64MB DDR 4ns (265/390 MHz) (31)
  26. ATI RADEON 7500 64MB DDR 4ns (270/460 MHz) (25)
  27. NVIDIA GeForce4 MX 440Ultra 64MB DDR 3.6ns (270/500 MHz) (22)
  28. ATI RADEON 8500 128MB DDR 3.3ns Samsung (275/550 MHz) (29)
  29. ATI RADEON 8500XT 128MB DDR 3.3ns (300/600 MHz) (30)
  30. NVIDIA GeForce4 Ti 4600 GS 128MB DDR 2.8ns (310/680 MHz) (28)
  31. ATI RADEON 7500 64MB DDR 4ns (230/333 MHz) (34)
  32. ATI RADEON 9000 128MB DDR 3.6ns (250/400 MHz) (27)
  33. ATI RADEON 7500 64MB DDR 4ns (290/460 MHz) (33)
  34. NVIDIA GeForce2 Ti 64MB DDR 5ns (250/400 MHz) (32)
  35. NVIDIA GeForce2 Ti VX 64MB DDR 5.5ns (225/366 MHz) (35)
  36. SIS Xabre200 64MB DDR 4ns (200/366 MHz) (37)
  37. NVIDIA GeForce4 MX 420 64MB SDR 6ns (250/166 MHz) (36)
  38. Matrox Parhelia 128MB DDR 3.3ns (220/550 MHz) (Retail)(38)
  39. PowerVR/STM KYRO II 64MB SDR 5ns (175/175 MHz) (39)
  40. PowerVR/STM KYRO II SE 64MB SDR 4.7ns (200/200 MHz) (40)
  41. NVIDIA GeForce2 MX400 64MB SDR 6ns (200/183 MHz) (42)

Middle-End usability rating

  1. NVIDIA GeForce4 Ti 4200 64MB DDR 4ns (250/513 MHz) (1)
  2. NVIDIA GeForce3 Ti500 64MB DDR 3.8ns (240/500 MHz) (12)
  3. ATI RADEON 8500LE 64MB DDR 3.6ns (250/500 MHz) (7)
  4. NVIDIA GeForce3 Ti 200 64MB DDR 5ns (175/400 MHz) (2)
  5. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/500 MHz) (24)
  6. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/444 MHz) (3)
  7. ATI RADEON 8500 64MB DDR 3.6ns (275/550 MHz) (21)
  8. ATI RADEON 9000 Pro 64MB DDR 3.6ns (275/550 MHz) (4)
  9. ATI RADEON 7500 64MB DDR 4ns (265/390 MHz) (23)
  10. SIS Xabre600 64MB DDR 4ns (275/550 MHz) (11)
  11. NVIDIA GeForce4 Ti 4400 128MB DDR 3.6ns (275/550 MHz) (9)
  12. ATI RADEON 8500LE 128MB DDR 3.6ns (250/500 MHz) (14)
  13. NVIDIA GeForce4 MX 440 64MB DDR 5ns (270/400 MHz) (16)
  14. SIS Xabre400 64MB DDR 3.5ns (250/500 MHz) (13)
  15. SIS Xabre400 128MB DDR 4ns (250/500 MHz) (18)
  16. SIS Xabre600 128MB DDR 4ns (275/550 MHz) (17)
  17. ATI RADEON 7500 64MB DDR 4ns (270/460 MHz) (10)
  18. ATI RADEON 8500LE 64MB DDR 5ns (250/366 MHz) (15)
  19. ATI RADEON 8500LE 64MB DDR 4ns (230/460 MHz) (8)
  20. ATI RADEON 9000 Pro 128MB DDR 3.6ns (275/550 MHz) (6)
  21. NVIDIA GeForce4 MX 460 64MB DDR 3.6ns (300/550 MHz) (26)
  22. ATI RADEON 7500 64MB DDR 4ns (230/333 MHz) (29)
  23. NVIDIA GeForce4 Ti 4200 128MB DDR 3.3ns (250/550 MHz) (19)
  24. ATI RADEON 8500LE 64MB DDR 5ns (183/366 MHz) (25)
  25. NVIDIA GeForce4 MX 440Ultra 64MB DDR 3.6ns (270/500 MHz) (20)
  26. NVIDIA GeForce2 Ti 64MB DDR 5ns (250/400 MHz) (27)
  27. ATI RADEON 8500 128MB DDR 3.6ns Hynix (275/550 MHz) (30)
  28. NVIDIA GeForce4 Ti4600 128MB DDR 2.8ns (300/650 MHz) (28)
  29. ATI RADEON 7500 64MB DDR 4ns (290/460 MHz) (32)
  30. ATI RADEON 9000 128MB DDR 3.6ns (250/400 MHz) (22)
  31. ATI RADEON 9700 128MB DDR 2.8ns (325/620 MHz) (-)
  32. NVIDIA GeForce2 Ti VX 64MB DDR 5.5ns (225/366 MHz) (31)
  33. NVIDIA GeForce4 MX 420 64MB SDR 6ns (250/166 MHz) (34)
  34. SIS Xabre200 64MB DDR 4ns (200/366 MHz) (33)
  35. ATI RADEON 8500 128MB DDR 3.3ns Samsung (275/550 MHz) (35)
  36. ATI RADEON 8500XT 128MB DDR 3.3ns (300/600 MHz) (36)
  37. PowerVR/STM KYRO II 64MB SDR 5ns (175/175 MHz) (38)
  38. NVIDIA GeForce4 Ti 4600 GS 128MB DDR 2.8ns (310/680 MHz) (37)
  39. PowerVR/STM KYRO II SE 64MB SDR 4.7ns (200/200 MHz) (41)
  40. NVIDIA GeForce2 MX400 64MB SDR 6ns (200/183 MHz) (42)
  41. Matrox Parhelia 128MB DDR 3.3ns (220/550 MHz) (Retail)(43)

Low-End usability rating

  1. NVIDIA GeForce4 Ti 4200 64MB DDR 4ns (250/513 MHz) (3)
  2. NVIDIA GeForce3 Ti 200 64MB DDR 5ns (175/400 MHz) (1)
  3. ATI RADEON 8500LE 64MB DDR 3.6ns (250/500 MHz) (12)
  4. ATI RADEON 7500 64MB DDR 4ns (265/390 MHz) (11)
  5. NVIDIA GeForce3 Ti500 64MB DDR 3.8ns (240/500 MHz) (21)
  6. NVIDIA GeForce4 MX 440 64MB DDR 5ns (270/400 MHz) (10)
  7. ATI RADEON 7500 64MB DDR 4ns (230/333 MHz) (20)
  8. ATI RADEON 8500 64MB DDR 3.6ns (275/550 MHz) (27)
  9. SIS Xabre400 64MB DDR 3.5ns (250/500 MHz) (6)
  10. ATI RADEON 7500 64MB DDR 4ns (270/460 MHz) (4)
  11. SIS Xabre600 64MB DDR 4ns (275/550 MHz) (9)
  12. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/444 MHz) (7)
  13. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/500 MHz) (34)
  14. ATI RADEON 9000 Pro 64MB DDR 3.6ns (275/550 MHz) (2)
  15. SIS Xabre400 128MB DDR 4ns (250/500 MHz) (14)
  16. SIS Xabre600 128MB DDR 4ns (275/550 MHz) (16)
  17. NVIDIA GeForce2 Ti 64MB DDR 5ns (250/400 MHz) (15)
  18. ATI RADEON 8500LE 64MB DDR 5ns (250/366 MHz) (18)
  19. ATI RADEON 8500LE 128MB DDR 3.6ns (250/500 MHz) (23)
  20. NVIDIA GeForce4 MX 460 64MB DDR 3.6ns (300/550 MHz) (25)
  21. NVIDIA GeForce4 MX 440Ultra 64MB DDR 3.6ns (270/500 MHz) (19)
  22. ATI RADEON 8500LE 64MB DDR 5ns (183/366 MHz) (22)
  23. ATI RADEON 9000 Pro 128MB DDR 3.6ns (275/550 MHz) (8)
  24. ATI RADEON 8500LE 64MB DDR 4ns (230/460 MHz) (13)
  25. NVIDIA GeForce4 MX 420 64MB SDR 6ns (250/166 MHz) (29)
  26. NVIDIA GeForce4 Ti 4400 128MB DDR 3.6ns (275/550 MHz) (26)
  27. SIS Xabre200 64MB DDR 4ns (200/366 MHz) (28)
  28. NVIDIA GeForce2 Ti VX 64MB DDR 5.5ns (225/366 MHz) (24)
  29. NVIDIA GeForce4 Ti 4200 128MB DDR 3.3ns (250/550 MHz) (30)
  30. ATI RADEON 7500 64MB DDR 4ns (290/460 MHz) (33)
  31. PowerVR/STM KYRO II 64MB SDR 5ns (175/175 MHz) (35)
  32. ATI RADEON 9000 128MB DDR 3.6ns (250/400 MHz) (17)
  33. ATI RADEON 8500 128MB DDR 3.6ns Hynix (275/550 MHz) (36)
  34. NVIDIA GeForce2 MX400 64MB SDR 6ns (200/183 MHz) (39)
  35. PowerVR/STM KYRO II SE 64MB SDR 4.7ns (200/200 MHz) (38)
  36. NVIDIA GeForce4 Ti4600 128MB DDR 2.8ns (300/650 MHz) (37)
  37. ATI RADEON 8500 128MB DDR 3.3ns Samsung (275/550 MHz) (40)
  38. ATI RADEON 9700 128MB DDR 2.8ns (325/620 MHz) (-)
  39. ATI RADEON 8500XT 128MB DDR 3.3ns (300/600 MHz) (41)
  40. NVIDIA GeForce4 Ti 4600 GS 128MB DDR 2.8ns (310/680 MHz) (42)
  41. Matrox Parhelia 128MB DDR 3.3ns (220/550 MHz) (Retail)(43)






"Prospect rating" (rating from the previous 3Digest is in parentheses):

  1. ATI RADEON 9700 128MB DDR 2.8ns (325/620 MHz) (-)
  2. NVIDIA GeForce4 Ti 4600 GS 128MB DDR 2.8ns (310/680 MHz) (1)
  3. NVIDIA GeForce4 Ti 4600 128MB DDR 2.8ns (300/650 MHz) (2)
  4. NVIDIA GeForce4 Ti 4400 128MB DDR 3.6ns (275/550 MHz) (3)
  5. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/500 MHz) (4)
  6. NVIDIA GeForce4 Ti 4200 128MB DDR 3.3ns (250/550 MHz) (5)
  7. NVIDIA GeForce4 Ti 4200 64MB DDR 4ns (250/513 MHz) (6)
  8. NVIDIA GeForce4 Ti 4200 128MB DDR 4ns (250/444 MHz) (7)
  9. Matrox Parhelia 128MB DDR 3.3ns (220/550 MHz) (Retail)(8)
  10. ATI RADEON 8500 128MB DDR 3.6ns Hynix (275/550 MHz) (10)
  11. ATI RADEON 8500XT 128MB DDR 3.3ns (300/600 MHz) (9)
  12. ATI RADEON 8500 128MB DDR 3.3ns Samsung (275/550 MHz) (11)
  13. ATI RADEON 8500LE 128MB DDR 4ns (250/500 MHz)(12)
  14. NVIDIA GeForce3 Ti 500 64MB DDR 3.8ns (240/500 MHz) (13)
  15. ATI RADEON 8500 64MB DDR 3.6ns (275/550 MHz) (14)
  16. ATI RADEON 8500LE 64MB DDR 3.6ns (250/500 MHz) (15)
  17. ATI RADEON 8500LE 64MB DDR 4ns (230/460 MHz)(16)
  18. ATI RADEON 9000 Pro 64MB DDR 3.6ns (275/550 MHz) (18)
  19. NVIDIA GeForce3 Ti200 64MB DDR 5ns (175/400 MHz) (19)
  20. ATI RADEON 9000 Pro 128MB DDR 3.6ns (275/550 MHz) (20)
  21. ATI RADEON 8500LE 64MB DDR 5ns (250/366 MHz) (21)
  22. ATI RADEON 8500LE 64MB DDR 5ns (183/366 MHz) (22)
  23. NVIDIA GeForce4 MX 460 64MB DDR 3.6ns (300/550 MHz) (23)
  24. ATI RADEON 9000 128MB DDR 3.6ns (250/400 MHz) (24)
  25. NVIDIA GeForce4 MX 440Ultra 64MB DDR 3.6ns (270/500 MHz) (25)
  26. ATI RADEON 7500 64MB DDR 4ns (290/460 MHz) (26)
  27. ATI RADEON 7500 64MB DDR 4ns (270/460 MHz) (27)
  28. NVIDIA GeForce4 MX 440 64MB DDR 5ns (270/400 MHz) (28)
  29. ATI RADEON 7500 64MB DDR 4ns (265/390 MHz) (29)
  30. SIS Xabre600 128MB DDR 4ns (275/550 MHz) (30)
  31. SIS Xabre400 128MB DDR 4ns (250/500 MHz) (31)
  32. SIS Xabre600 64MB DDR 4ns (275/550 MHz) (32)
  33. ATI RADEON 7500 64MB DDR 4ns (230/333 MHz) (33)
  34. SIS Xabre400 64MB DDR 3.5ns (250/500 MHz) (34)
  35. NVIDIA GeForce4 MX 420 64MB SDR 6ns (250/166 MHz) (35)
  36. NVIDIA GeForce2 Ti 64MB DDR 5ns (250/400 MHz) (36)
  37. NVIDIA GeForce2 Ti VX 64MB DDR 5.5ns (225/366 MHz) (37)
  38. SIS Xabre200 64MB DDR 4ns (200/366 MHz) (38)
  39. NVIDIA GeForce2 MX400 64MB SDR 6ns (200/183 MHz) (39)
  40. PowerVR/STM KYRO II SE 64MB SDR 4.7ns (200/200 MHz) (40)
  41. PowerVR/STM KYRO II 64MB SDR 5ns (175/175 MHz) (42)

P.S. We understand that such way of making ratings is far from perfect, therefore we await suggestions and comments from our readers to be considered in our next 3Digest.

We also await any remarks and recommendations from readers about improving the 3Digest itself.

9. Screenshot gallery (Windows XP)

All test screenshots are compiled into a gallery where you can qualify the work of a chosen videocard and compare it to the reference images taken on NVIDIA GeForce3. All shots were taken at the same 1024x768 resolution at the maximum quality.

All authors:

Evgeniy Zaycev



Alexander Kondakov



Andrey Vorobiev



Write a comment below. No registration needed!


Article navigation:



blog comments powered by Disqus

  Most Popular Reviews More    RSS  

AMD Phenom II X4 955, Phenom II X4 960T, Phenom II X6 1075T, and Intel Pentium G2120, Core i3-3220, Core i5-3330 Processors

Comparing old, cheap solutions from AMD with new, budget offerings from Intel.
February 1, 2013 · Processor Roundups

Inno3D GeForce GTX 670 iChill, Inno3D GeForce GTX 660 Ti Graphics Cards

A couple of mid-range adapters with original cooling systems.
January 30, 2013 · Video cards: NVIDIA GPUs

Creative Sound Blaster X-Fi Surround 5.1

An external X-Fi solution in tests.
September 9, 2008 · Sound Cards

AMD FX-8350 Processor

The first worthwhile Piledriver CPU.
September 11, 2012 · Processors: AMD

Consumed Power, Energy Consumption: Ivy Bridge vs. Sandy Bridge

Trying out the new method.
September 18, 2012 · Processors: Intel
  Latest Reviews More    RSS  

i3DSpeed, September 2013

Retested all graphics cards with the new drivers.
Oct 18, 2013 · 3Digests

i3DSpeed, August 2013

Added new benchmarks: BioShock Infinite and Metro: Last Light.
Sep 06, 2013 · 3Digests

i3DSpeed, July 2013

Added the test results of NVIDIA GeForce GTX 760 and AMD Radeon HD 7730.
Aug 05, 2013 · 3Digests

Gainward GeForce GTX 650 Ti BOOST 2GB Golden Sample Graphics Card

An excellent hybrid of GeForce GTX 650 Ti and GeForce GTX 660.
Jun 24, 2013 · Video cards: NVIDIA GPUs

i3DSpeed, May 2013

Added the test results of NVIDIA GeForce GTX 770/780.
Jun 03, 2013 · 3Digests
  Latest News More    RSS  

Platform  ·  Video  ·  Multimedia  ·  Mobile  ·  Other  ||  About us & Privacy policy  ·  Twitter  ·  Facebook


Copyright © Byrds Research & Publishing, Ltd., 1997–2011. All rights reserved.