Here is a table to compare some of the major features.
|Memory||Cooling||Video||Software DVD player|
|AOpen PA256 Pro||32 MBs of 6ns Infineon DDR at 333 MHz||Standard||Video-out||No|
|ELSA Gladiac||32 MBs of 6ns Infineon DDR at 333 MHz||Standard||Optional||Yes *|
|Leadtek WFGF2||32 MBs of 6ns Infineon DDR at 333 MHz||Excellent||Video-out||Yes|
* Not in retail box but can be requested through technical support.
Performance - Test Setup
|CPU||Intel PIII 866|
|Motherboard (BIOS rev.)||OR840Intel OR840 (BIOS OR840700.86E.0219.803.0003081415)|
|Memory||2 Modules of 128 MB 800 MHz RDRAM|
|CPU||Intel PIII 600E @ 800 MHz|
|Motherboard (BIOS rev.)||MS BX Master (BIOS 5.1 031400)|
|Memory||1 Module of 128 MB Crucial PC133 CAS2 SDRAM|
|AOpen PA256 Pro||4.12.01.0516|
|Reference GeForce 256 DDR||4.12.01.0516|
|OS Version||Windows 98 SE 4.10.2222 A|
|Quake 3 Arena||Retail versioncommand line = +set cd_nocd 1 +set s_initsound 0|
|Shogo||V2.14Advanced Settings = disable sound, disable music, disable movies, disable joysticks,enable optimized surfaces, enable triple buffering, enable single-pass multi-texturingHigh Detail Settings = enabledFortress Demo|
|3DMark 2000||16-bit settings = 16 bit textures, 16-bit Z-buffer, triple buffering32-bit settings = 32-bit textures, 24-bit Z-buffer, triple bufferingAll test were done with high detail only.|
Seeing as most of the boards are based on the NVIDIA 5.16 driver, I don't expect to see much of a performance difference. Aside from vsync I've left all the default settings alone to represent what most people will see. This means fast writes are on and no other changes were made to tweak the visuals. I did also compare default settings of each manufacturer and found no major differences. Layout differences typically give more stability benefit than raw performance unless there is a major change to the design. I expect our scores to be very close among the competing boards.
Test Results - Shogo
At this low resolution the benchmark is much more CPU dependant although we do see the GeForce2 based cards pull slightly ahead of the older GeForce reference card.
Lifting the resolution upwards begins to show the fill-rate advantage of the GeForce2 based cards. All the GeForce2 cards are basically tied.
At this high resolution and in 16-bit color the GeForce2 boards take a nice lead over the older GeForce board but still score right next to each other.