iXBT Labs - Computer Hardware in Detail

Platform

Video

Multimedia

Mobile

Other

Hercules 3D Prophet 9800 XT 256MB Video Card Review

November 17, 2003






CONTENTS

  1. Video cards features
  2. Testbed, test tools, 2D quality
  3. Test results: Quake3 ARENA
  4. Test results: Serious Sam: The Second Encounter 
  5. Test results: Return to Castle Wolfenstein
  6. Test results: Code Creatures DEMO
  7. Test results: Unreal Tournament 2003 
  8. Test results: Unreal II: The Awakening
  9. Test results: RightMark 3D
  10. Test results: TRAOD: Benchmarking in paris5_4
  11. Test results: TRAOD: Benchmarking in paris1c
  12. Test results: TRAOD: Benchmarking in paris2g
  13. Test results: HALO: Combat Evolved 
  14. Test results: Splinter Cell
  15. Quality issues (in other tests)
  16. Conclusion


We have recently published the analytical research that outlines the price niches in the 3D GPU sector. It shows that the High-End products take almost 0.1% of the whole video card market in Russia (it makes 7-9% in Europe and USA because of the higher standard of life and lower prices).

I've already expressed what I think about the pricing. High-End products have become too expensive, and they only morally hold the gaming 3D crown.

Are such products priced at $500 worth our attention? Well, all new technologies are tested on such cards that is why it's better to describe them using powerful accelerators. Besides, if users pay such sum of money they have the right to get as detailed information on performance of such cards as possible because in most cases all such functions (like AA, anisotropy) are going to be enabled. Owners of Low-end and some Middle-end cards hardly need such functions because of a relatively low speed. After all, most low-end cards are actually designed only for 2D and they just optionally offer 3D functions. That is why we carefully test every High-End product. We have sifted a lot of cards of such class (some of them stepped down to the Middle-end sector, some became dated, some simply died). Some of the cards listed below sometimes crowned the 3D top.

Theoretical materials and reviews of video cards which concern functional properties of the GPU ATI RADEON 9500/9700/9800

However, it doesn't turn us away from problems of the low- and middle-end sectors. Some of such cards are highlighted in our 3Digest, others are reviewed separately. Today it's high time to make a kind of a general review like that we made for sub-$100 cards. I hope by the New Year we will manage to cover all sub-$80 accelerators in one review.

And today we are testing one more High-End card based on the RADEON 9800 XT and made by Hercules. This is Guillemot's subdepartment that produces all cards under the 3D Prophet trade mark on ATI's GPUs. 

This RADEON 9800 XT based card is almost the fastest High-End solution from ATI Technologies camp because...

Well, let's see why :)

Card

Hercules 3D Prophet 9800 XT 256MB



 
Hercules 3D Prophet 9800 XT 256MB
The card has AGP x8/x4 interface, 256MB DDR SDRAM in 8 chips on both PCB sides.

Hynix 2.5ns memory chips, HY5DU573222 (corresponds to 400 (800) MHz), memory clocked at 365 (730) MHz, GPU at 412 MHz. 256-bit memory bus.





 
Comparison with the reference design, front view
Hercules 3D Prophet 9800 XT 256MB Reference card ATI RADEON 9800 XT







 
Comparison with the reference design, back view
Hercules 3D Prophet 9800 XT 256MB Reference card ATI RADEON 9800 XT







 

This is obviously a copy of the reference card. Although earlier all RADEON 9800 PRO based cards were made at the same factory, released with ATI's part number and then sent to ATI's partners, Hercules's cards had their own peculiarities (blue textolite, copper coolers on GPU and memory chips). Now all RADEON 9800 XT based cards except those for ASUS are produced at the same factory, and the partners only attach their stickers.

The cooler was already discussed in the basic RADEON 9800 XT review. Let me just briefly repeat it. 
 

Hercules 3D Prophet 9800 XT 256MB
This beautiful cooler is also very efficient. Moreover, its dimensions let the card take only one AGP slot.

On the back it has a copper plate that is reliably pressed against the memory chips with two buckles (both sinks are screwed to the card). A special jut on the plate is attached to the PCB in the place where the core is located.





Remember that the cooler takes only one slot, and the fist PCI slot is free and doesn't hamper to cool the card:


 

The fan is pretty quiet (but it starts buzzing when the card overheats). Some owners of the 9800XT based cards say that the reference system is not that quiet but we didn't come across such cards.

The card supports VIVO provided that the Rage Theater 200 chip is installed (there is space for for it).

Have a look at the box and accessories. 
 

Hercules 3D Prophet 9800 XT
The box is made of thick cardboard and traditionally pictures a monster. The clock speeds are indicated as well.






 
 
Hercules 3D Prophet 9800 XT
The box contains a user guide, software CD (including Hercules overclocker), PowerDVD, coupon for Half-Life2, TV extension cords, DVI-to-d-Sub and SVideo-to-RCA adapters, external power supply adapter.



 

Testbed and drivers

Testbed: 

  • Pentium 4 3200 MHz based computer:
    • Intel Pentium 4 3200 MHz CPU;
    • DFI LANParty Pro875 (i875P) mainboard; 
    • 1024 MB DDR SDRAM; 
    • Seagate Barracuda IV 40GB HDD; 
    • Windows XP SP1; DirectX 9.0b;
    • ViewSonic P810 (21") and ViewSonic P817 (21") monitors.
    • ATI drivers v6.396 (CATALYST 3.9).

VSync off, S3TC off in applications. 

Video cards for comparison:

  • ATI RADEON 9800 XT (412/365 (730) MHz, 256 MB DDR).
  • NVIDIA GeForce Fx 5950 Ultra (475/475 (950) MHz, 256 MB DDR, driver 52.16).

Let me explain you why this card is considered the most powerful one among RADEON 9800 XT based ones. The matter is that starting from the CATALYST version 3.8 such cards possess the OverDrive function which allows increasing the GPU clock at certain temperatures.




On all the RADEON 9800 XT cards tested the clock speed lifted mainly up to 418 MHz and sometimes up to 432MHz. On the Hercules card it jumped up to 445(!) MHz just occasionally dropping down to 432 MHz. The RivaTuner proves it:




When overclocked, the card reached 470/760 MHz. 

Test results

Before we start examining 2D quality, I should say there are no complete techniques for objective 2D quality estimation because: 

  1. 2D quality much depends on certain samples for almost all modern 3D accelerators; 
  2. Besides videocards, 2D quality depends on monitors and cables; 
  3. Moreover, certain monitors might not work properly with certain video cards. 

With the ViewSonic P817 monitor and BNC Bargo cable the cards showed excellent quality at the following resolutions and clock speeds: 

Hercules 3D Prophet 9800 XT 256MB  1600x1200x85Hz, 1280x1024x120Hz, 1024x768x160Hz

 

Test results: performance

Conventional signs: ANISO 8xP - Anisotropic 8x Performance (earlier it was called Balanced), ANISO 8xQ - Anisotropic 8x Quality, ANISO 16xQ - Anisotropic 16x Quality. 

Some time ago we decided not to compare anymore ATI's maximum anisotropic quality of 16x to two NVIDIA's modes. The ANISO 8x Quality mode delivered the real maximum quality with both trilinear filtering and anisotropy working to their full capacity. The ATI 16x Quality showed sharper images due to the 16th degree but on some surfaces the filtering quality was lower. That's the way ATI's anisotropy works. That is why we thought it was more correct to compare this ATI's mode with NVIDIA's Performance and Quality. 

But NVIDIA's optimization policy changes the things and we do not know anymore if there are applications where NVIDIA's anisotropy works to its full capacity. That is why we consider that it's correct to compare ANISO 16xQ (ATI) to ANISO 8xQ (NV). Both have their strong and weak points, but in general they compensate each other. 

Test applications: 

  • Return to Castle Wolfenstein (MultiPlayer) (id Software/Activision) - OpenGL, multitexturing, ixbt0703-demo, test settings - maximum, S3TC OFF, the configurations can be downloaded from here 
  • Serious Sam: The Second Encounter v.1.05 (Croteam/GodGames) - OpenGL, multitexturing, ixbt0703 demo, test settings: quality, S3TC OFF 
  • Quake3 Arena v.1.17 (id Software/Activision) - OpenGL, multitexturing, ixbt0703 demo, test settings - maximum: detailing level - High, texture detailing level - #4, S3TC OFF, smoothness of curves is much increased through variables r_subdivisions "1" and r_lodCurveError "30000" (at default r_lodCurveError is 250 !), the configurations can be downloaded from here 
  • Unreal Tournament 2003 v.2225 (Digital Extreme/Epic Games) - Direct3D, Vertex Shaders, Hardware T&L, Dot3, cube texturing, default quality 
  • Code Creatures Benchmark Pro (CodeCult) - the game that demonstrates card's operation in DirectX 8.1, Shaders, HW T&L. 
  • Unreal II: The Awakening (Legend Ent./Epic Games) - Direct3D, Vertex Shaders, Hardware T&L, Dot3, cube texturing, default quality 
  • RightMark 3D v.0.4 (one of the test scenes) - DirectX 8.1, Dot3, cube texturing, shadow buffers, vertex and pixel shaders (1.1, 1.4). 
     
  • Tomb Raider: Angel of Darkness v.49 (Core Design/Eldos Software) - DirectX 9.0, three demo scenes, maximum quality, Depth of Fields PS20 off. The settings are equal for all the cards tested.




If you need patch 49 which is not easy to find and the demo benchmarks let me know by email. 
 

  • HALO: Combat Evolved (Microsoft) - Direct3D, Vertex/Pixel Shaders 1.1/2.0, Hardware T&L, high quality

  •  
  • Tom Clancy's Splinter Cell v.1.2b (UbiSoft) - Direct3D, Vertex/Pixel Shaders 1.1/2.0, Hardware T&L, Very High Quality; demo 1_1_2_Tbilisi

  •  

    Apart from testing the performance we'll also check quality of the RADEON 9800 XT and GeForce FX 5950 Ultra based cards in some benchmarks. First of all, the rumor has it that ATI's drivers have some cheats that reduce quality, and secondly, the developers released a new driver version (3.9) which should be tested as well. But it will be thoroughly tested in our 3Digest. 
     

    Quake3 Arena
















    Light modes without AA and anisotropy: parity

    AA enabled: the card loses in 1024x768 but takes the lead in 1600x1200

    Anisotropy enabled: Hercules R9800XT loses the game

    AA & anisotropy enabled: it's different
     
     

    Serious Sam: The Second Encounter
















    Light modes without AA and anisotropy: loses

    AA enabled: loses

    Anisotropy enabled: parity

    AA & anisotropy enabled: loses
     
     

    Return to Castle Wolfenstein (Multiplayer)
















    Light modes without AA and anisotropy: parity

    AA enabled: it goes on a par and sometimes falls a little behind 

    Anisotropy enabled: Hercules wins

    AA & anisotropy enabled: the same
     
     

    Code Creatures
















    Light modes without AA and anisotropy: a bit behind

    AA enabled: fiasco

    Anisotropy enabled: victory

    AA & anisotropy enabled: Hercules 9800 XT goes faster than the others but this mode is hardly playable
     
     

    Unreal Tournament 2003
















    Light modes without AA and anisotropy: Hercules wins the battle (though it's generally parity)

    AA enabled: victory

    Anisotropy enabled: victory

    AA & anisotropy enabled: victory

    Quality (Attention! The large screenshots are made in BMP and RAR packed, each measuring over 2 MB)
     

    RADEON 9800 FX 5900
    Example 1
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    Example 2
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    Example 3
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences





    The bit-per-bit comparison shows the difference only for moving objects (sky, weapon, shadows). I didn't resort to the AutoContrast not to distinguish the pixels that have slight color difference. Gamers won't notice it. 

    I don't know how the guys from NVIDIA managed to reveal such difference between the RADEON and GeForce at their conference, but our benchmarks don't show such a great difference. By the way, earlier we analyzed quality in this test and found out that NVIDIA simplified the trilinear filtering in some modes. 

    I don't understand why NVIDIA used exactly this benchmark to discredit ATI if they are not stainless themselves?
     
     

    Unreal II: The Awakening
















    The Hercules card wins in all modes.

    Quality (Attention! The large screenshots are made in BMP and RAR packed, each measuring over 2 MB)
     

    RADEON 9800 FX 5900
    Example 1
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    Example 2
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences





    Again, there is no much difference except for the moving objects.
     
     

    RightMark 3D
















    Well, this test uses a lot of shaders, and the Hercules R9800XT easily wins the battle.
     
     

    TR:AoD, Paris5_4 DEMO







    The scenes with a great number of shaders make the RADEON 9800 XT a victor.
     
     

    TR:AoD, Paris1c DEMO







    The same.
     
     

    TR:AoD, Paris2g DEMO







    This scene uses much fewer shaders that is why the NV38 comes very close to the 9800XT, though the is still faster.
     
     

    HALO: Combat Evolved










    Light modes without AA and anisotropy: victory

    This game doesn't support AA.

    Anisotropy enabled: Hercules R9800XT win this round!
     
     

    Splinter Cell










    Light modes without AA and anisotropy: parity.

    AA is not supported. 

    Anisotropy enabled: Hercules is a little ahead.
     
     

    Quality

    As you can see, the competitors don't differ much in quality. Let's also have a look at the game the shader quality in which much depends on the driver version. I mean the Morrowind.

    Attention! The large screenshots are made in BMP and RAR packed, each measuring over 2 MB.
     

    RADEON 9800 FX 5900
    Morrowind, Example 1
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    Morrowind, Example 2
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences




    There's difference again only for moving objects (well, it's impossible to make a screenshot at the same moment). The screenshots are also a response to those who blame ATI's anisotropy on sloping surfaces: its anisotropy doesn't differ from NVIDIA's "true" one.

    The last comes the AquaMark3. 
     

    RADEON 9800 FX 5900
    AquaMark3, Example 1
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    AquaMark3, Example 2
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    AquaMark3, Example 3
    No AA, No ANISO






    Differences



    AA4x, ANISO 16xQ/8xQ






    Differences



    The mist made from small polygons and the multipoligonal explosion that demonstrates Overdraw make ATI look inferior. What's that?

    NVIDIA's fans publicly point at the shortcomings in ATI's realization, and do it primarily at seminars and conferences.

    The example of the UT2003 demonstrates that if a company has problems with a certain benchmark, it also indicates its competitors' flaws in the same test. In the AquaMark3 the RADEON has an evident defect. On the other hand, ATI assures that there are no cheats in this benchmark, and this is the programmers who are to be blamed (though I don't understand whose programmers are faulty: either the AquaMark3 has some peculiar ways of realization of the DX functions, or the programmers at ATI use their own ones and the AquaMark3 doesn't detect it).

    ATI seems to have forgotten about quality and keeps on making mistakes. If it really is just a mistake in the AquaMark3, they will correct it soon, because cheats usually live forever... But if it's a cheat, it will soon become clear that ATI doesn't play a fair game, like NVIDIA. 

    Conclusion

    The Hercules RADEON 9800 XT graphics card is the fastest accelerator today provided that all such cards have the clock speed of 445 MHz (i.e. not only this sample).

    If you are ready to pay $500 for a 3D accelerator the Hercules 3D Prophet 9800 XT can become the best choice (provided that the price is not too higher compared to other 9800 XT based cards).

    Its quality has both advantages and disadvantages. On the one hand, the RADEON 9800 XT never loses and even sometimes looks superior (for example, in Yager). On the other hand, the quality degrades in some benchmarks, and it happens mostly through the programmers' fault.

    In our 3Digest you can find full comparison characteristics for video cards of this and other classes.
     
     

    Andrey Vorobiev (anvakams@ixbt.com
     

    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.