As the clash of the Titans continues between NVIDIA and chief-rival ATI the battle for the mid and low price point continues to bear down on consumer like a hangover after a frat party. Unlike the pain you experience the following day, however, this GPU hangover yields great benefits to the consumers by way of a competitive market place and lower prices.

ATI’s 1950 series GPU is a fully DirectX 9 compliant part that supports Shader Model 3.0 and comes in a variety of configurations and price points. The two we’re looking at in this round-up are the XT and Pro variations.

Radeon 1950XT Features:
• 384 million transistors on 90nm fabrication process
• Up to 48 pixel shader processors
• 8 vertex shader processors
• Up to 256-bit 8-channel GDDR4 memory interface
• Native PCI Express x16 bus interface

CrossFire
• Multi-GPU technology
• Four modes of operation:
• Alternate Frame Rendering (maximum performance)
• Supertiling (optimal load-balancing)
• Scissor (compatibility)
• Super AA 8x/10x/12x/14x (maximum image quality)

Radeon 1950Pro Features:
• 36 pixel shader processors
• 8 vertex shader processors
• Up to 256-bit 8 channel GDDR3 memory interface
• Native PCI Express x16 bus interface
• Plug-and-play (native) CrossFire

CrossFire
• Multi-GPU technology
• Four modes of operation:
• Alternate frame rendering for maximum performance
• Supertiling for optimal load-balancing
• Scissoring for compatibility
• Super AA for maximum image quality
• Native CrossFire support simplifies setup by requiring no dedicated slave or master hardware
• 24-bit CrossFire connection enables high resolutions and refresh rates
• Supports the broadest range of platforms for both Intel and AMD

Common Features for both XT and Pro:

Ring Bus Memory Controller
• Fully associative texture, color, and Z/stencil cache designs
• Hierarchical Z-buffer with Early Z test
• Lossless Z Compression (up to 48:1)
• Fast Z-Buffer Clear
• Optimized for performance at high display resolutions, including widescreen HDTV resolutions

Ultra-Threaded Shader Engine
• Support for Microsoft DirectX 9.0 Shader Model 3.0 programmable vertex and pixel shaders in hardware
• Full speed 128-bit floating point processing for all shader operations
• Up to 512 simultaneous pixel threads
• Dedicated branch execution units for high performance dynamic branching and flow control
• Dedicated texture address units for improved efficiency
• 3Dc+ texture compression o High quality 4:1 compression for normal maps and two-channel data formats
• High quality 2:1 compression for luminance maps and single-channel data formats
• Complete feature set also supported in OpenGL 2.0

---