“Both NVIDIA and ATI graphics processors are based on a unified architecture with programmable stream processors that can handle not only traditional graphics (vertex, geometry, texture) processing but also general purpose computing operations. However, NVIDIA have an ace up their sleeves that currently gives them an edge over ATI and that ace is called PhysX.
PhysX is NVIDIA’s proprietary technology. Only NVIDIA graphics cards are capable of accelerating games and applications that support PhysX. This gives them an important leverage over their competitor, ATI, who has made a remarkable comeback on the back of their Radeon HD 4000 series of graphics cards. As popular as the current ATI graphics cards, they cannot accelerate games and applications that support PhysX.”
Here are some more Graphics Card articles from around the web:
- GeForce GTX 295 SLI: Super Heavy Weights Championship. Final Round @ X-bit Labs
- Inno3D GTS 250 TwinTurbo Pro – Making the most of G92 @ Guru of 3D
- nVidia GeForce GTX 280 @ LostCircuits
- Windows 7 (Final OEM) vs. Windows Vista VGA Performance @ Tweaktown
- Sapphire Radeon HD 4890 2GB Vapor-X @ Tweaktown
- Powercolor Radeon HD 4730 PCS @ Ultimate Hardware
- Palit Radeon 4870 Sonic Dual Edition in CrossfireX @ Techware Labs
Source: Tech ARP
In one corner is the CPU calculated physics software called Havok and in the other corner is what used to be Ageia’s and is now nVIDIA’s proprietary GPU physics program, PhsyX. TechARP examines both solutions from various standpoints. As a game developer both accelerators have benefits and drawbacks to programming and to the gamer, if no one is programming one or the other then they will never appear in games. More interesting is the examination of the performance of the two solutions in physics accelerated games. Take a look at the current leader here.