iXBT Labs - Computer Hardware in Detail






Prolink PixelView GeForce FX 5900 128MB
Benchmarking in Tomb Raider: Angel of Darkness

September 30, 2003


  1. Video card's features
  2. Testbed, test tools, 2D quality
  3. Test results: TRAOD: Benchmarking in paris5_4
  4. Test results: TRAOD: Benchmarking in paris1c
  5. Test results: TRAOD: Benchmarking in paris2g
  6. Conclusion

Today we will test one more video card based on the NVIDIA GeForce FX 5900. It seems that there's nothing to add to this subject. But it's wrong. First of all, High-End products and any alterations in this price niche indicate that the developers work hard trying to improve performance and drivers.

Secondly, performance of expensive video cards is always in the center of attention, especially of those who can't afford them, and every mistake of driver developers is strictly judged. Today we'll deal with a new (in our lab, not on the market) card based on the GeForce FX 5900 and test performance of such cards in the new sensational benchmark based on the game Tomb Raider: Angel of Darkness. Although NVIDIA's fans should enjoy the whole process, it's really only Prolink's card that gives real pleasure.

Theoretical materials and reviews of video cards which concern functional properties of the GPU NVIDIA GeForce FX

As usual, Prolink, Taiwanese manufacturer, offers a full line of video cards based on NVIDIA's processor. By the way, there are also also TV tuners and other multimedia devices produced under the PixelView trade mark.

The advantage of this FX 5900 based card is the lowest price on our market, though the card combines the reference design, high-grade quality and gold-plated surface...


Prolink PixelView GeForce FX 5900 128MB

Prolink PixelView GeForce FX 5900 128MB
AGP x8/x4/x2 interface, 128 MB DDR SDRAM memory in 8 chips on top.

Hynix 2.2ns memory chips, it corresponds to 454 (908) MHz but the memory works at 426 (850) MHz. GPU clocked at 400 MHz. 256bit memory bus.

Comparison with the reference design, front view
Prolink PixelView GeForce FX 5900 128MB Reference card NVIDIA GeForce FX 5900

Comparison with the reference design, back view
Prolink PixelView GeForce FX 5900 128MB Reference card NVIDIA GeForce FX 5900

The card is entirely based on the reference design. Plus, it's covered with evaporated gold:

If you remember, Triplex cards are also covered with an evaporated layer, but a silver one. But the quality of evaporation differs: the silver looks thicker and all components are carefully isolated to avoid electrical collisions. The approach in case of Prolink's card is different, but the company assures that it's not a fiction or fake, and this technology should protect the card from pickups and make for better cooling.

But the main aim of such gilding is a marketing one: it's a beautiful product, looks pretty nice and you can boast to your friends about such a golden card. :-)

Have a look at the cooler.

Prolink PixelView GeForce FX 5900 128MB

There is a single copper cooler for the GPU and memory chips. The fan is shifted to the left and serves as a turbine that draws air through the sink.

The cooler is a bit noisy. The noise is not unpleasant or too loud but those who prefer quiet PC cases are not going to like it.

Here's the GPU itself:

It's made in May 2003.

Note that the video card supports VIVO via the Philips 7108:

Prolink PixelView GeForce FX 5900 128MB

The card ships together with a user manual, software CD (WinDVD included), VIVO adapter/splitter, TV extension cords, S-Video-to-RCA adapter, external power supply splitter.

Prolink PixelView GeForce FX 5900 128MB

The package matches the card. The tender silvery box with buttons and a tape has the card name and basic features written on it.

Testbed and drivers


  • 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 9.0a;
    • ViewSonic P810 (21") and ViewSonic P817 (21") monitors.
    • NVIDIA drivers v51.75.

VSync off, S3TC off in applications.

Cards used for comparison:

  • Hercules 3D Prophet 9800 PRO (RADEON 9800 PRO, 380/340(680) MHz, 128 MB DDR, driver 6.378).
  • Sapphire Atlantis RADEON 9800 (325/290(580)MHz, 128 MB DDR, driver 6.378).

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 card showed excellent quality at the following resolutions and clock speeds:

Prolink PixelView GeForce FX 5900 128MB 1600x1200x85Hz, 1280x1024x120Hz, 1024x768x160Hz

Cards' performance

Test application:

  • Tomb Raider: Angel of Darkness v.49 (Core Design/Eldos Software) - DirectX 9.0, three demo scenes, test settings shown below.

If you want to get the demo benchmarks that we use let me know by e-mail.

First the cards were tested with the maximum settings, just Depth of Fields PS20 and PS20 Shadows were turned off (the first one worked with errors on the GeForce FX with patch 49, and in the second case the game failed to start on the GeForce FX again on patch 49). These bugs are fixed on the patch 52, but it's not any case to benchmark with patch 52.

All settings were the same for all the cards tested.

The settings circled were originally changed relative to the default ones. But we got NVIDIA's official respond where they insisted on testing with the default settings:

It should be noted that the the game and patch 49 were reinstalled for testing cards from a different camp to avoid problems with game settings for the RADEON 9800 PRO and GeForce FX 5900. Also note that when the game detected the RADEON 9800 PRO and the GeForce FX 5900 the default settings were also equal and they are shown above.

As you can see, some functions are disabled, and the Cubemaps reflections section sets 16bit color and Z buffer depths.

Paris5_4 DEMO

This scene intensively uses shader technologies.

What can I say about the GeForce FX 5900? Unfortunately, nothing.

Well, when the DX9 game was released the low pixel shader speed couldn't assure any decent performance. It's due time to cut prices for the FX5900! I hope NVIDIA's price policy will rescue such cards as the number of games that actively use PS 2.0 will be increasing. Note that disabling Glow, FOG and trilinear filtering has almost the same effect on both cards.

Quality (all quality parameters were measured at the maximum settings):

RADEON 9800/FX 5900 without DoF PS2.0 FX 5900 with DoF PS2.0 (Patch 49!)
Example 1

Example 2

Example 3

The Depth of Fields PS2.0 (DoF PS2.0) function (PATCH 49 only!) works the wrong way in case of the GeForce FX as it blurs the whole image instead of just background like in case of the RADEONs. When the DoF is disabled the RADEON 9800 and the GeForce FX have the same image quality. Besides, the PS 2.0 Shadows work flawlessly on the RADEON 9800 and buggy on the FX 5900 (the left screenshots show the picture obtained with the RADEON 9800; the FX5900 shows the same picture but without the hero's shadow). Sometimes you can notice bugs in both cases but it's the flaw in the game itself.

Attention! These promblems are only on patch 49. Reader can compare both results.

Paris1c DEMO

Obviously, this picture is still sad for the GeForce FX though the difference is not that big. No shaders can raise the speed if the chip itself is not able of it

This scene uses shader technologies to a less extent. The maximum settings affect the GeForce FX greater than the RADEON 9800 PRO (the latter hardly changes its speed at all, sometimes it works even slower with the default settings).


RADEON 9800/FX 5900 without DoF FX 5900 with DoF (Patch 49!)
Example 1

Example 2

Example 3

Paris2g DEMO

The scene that doesn't use so many shader technologies doesn't let the FX 5900 lose so awfully (and with the default settings the GFFX and R9800 PRO have equal scores). Well, the more shaders, the weaker the GeForce FX.

Strangely enough, but the RADEON 9800 PRO works faster at the maximum settings in this game than at the default ones. We checked it several times, that is why there must be a flaw in the game itself.


RADEON 9800/FX 5900 without DoF PS2.0 FX 5900 with DoF PS2.0 (Patch 49!)
Example 1

Example 2

There are no test results in the AA and anisotropy modes as the FPS is low even without them. Just remember that they do not change the general picture of the cards' performance.


Just before this review was published we found the patch 52 for this game on the Net. It doesn't support benchmarking anymore. But it's still possible to compare the GeForce FX 5900 and RADEON 9800 PRO with the FRAPS. The speeds remain the same, that is why the patch 52 doesn't improve performance of the NV35.

Remember! Patch 52 fixed many bugs (DoF PS2.0, PS2.0 Shadows), but it has not benchmark.

Also note that we checked all measurements taken with the integrated benchmark with the FRAPS utility. The figures differed by just a little margin, that is why we can consider the scores to be true. It was done to prevent claims that the integrated benchmark works incorrectly.


The card itself looks pretty good among its competitors: it has a low price and such advantages as perfect build quality, attractive appearance and package.

On the other hand (it concerns all GeForce FX based cards), such cards fall far behind their RADEON based competitors in the games that use a lot of DirectX 9.0 shader technologies. As to the TR:AoD game, it's for a user to choose settings. I didn't notice much difference between maximum and default settings. Meticulous users will notice it. That is why we tested both modes. I didn't compare screenshots with the maximum and default settings since I want to wait for the official release of NVIDIA's drivers v5x.xx, and secondly, benchmarking in TR:AoD will be soon carried out once again, and we will closer exmine the difference in quality.

As to the GeForce FX 5900, there is no wayouts other than to cut prices for the FX5900 relative to the RADEON 9800 PRO to increase their demand in spite of a low shader speed.

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

In closing let me tell you a little on the program called "The Way It's meant To Be Played". You have probably noticed such logo on some game CDs and packages of graphics cards from various card makers. It means that NVIDIA jointly works with the game developers to provide best compatibility of their products with NVIDIA based cards. NVIDIA's official position is as follows: This is just a program of ensuring compatibility with NVIDIA's products, and it's not aimed at all at accelerating NVIDIA's GPUs and slowing down competitors' ones.

Recently we got the first Russian game within the frames of this program - Operation: Silent Storm from NIVAL (Publisher 1Ñ).

Above you can see NVIDIA's official position regarding the program "The Way It's meant To Be Played". You can get more information about it on Tom's Hardware site.

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.