A week or so ago we had the opportunity to the Sapphire HD3850 video card running as a single card and then running a pair of them in Crossfire configuration. What we were able to determine after all of our tests was that under idea condition the two video cards running side by side made a substantial difference in our scores.
Today we are going to be looking at the big boy in the AMD/ATI house the HD3870 X2, if you were to give this card a quick glance you might just say it was one of the long red circuit boards AMD/ATI have become famous for. It would not be until you turned the graphics card over and you saw that the card had a pair of GPUs attached to it would you realize that there was something special here.
At the heart of the Sapphire HD3870 X2 graphics card is the R680 graphics processing unit which consists of two RV670 GPU cores (two Sapphire HD3780). The raw power of the two GPU promises to provide you with a gaming experience like none you have ever experienced prior to this. You are now able to enjoy the Crossfire experience we showed you a week ago using a single video card and over 1 teraFLOP of the GPU Computer Power in a single card solution. The card is available in two different sizes, 512 and 1 GB, with support for DirectX 10, Shader Model 4 and OpenGL acceleration and support for Windows Vista.
Radeon
HD 2600 XT
Radeon
HD 3850
Radeon
HD 3870
ROPs
120
320
320
Transistors
4 x2
16
16
Shader units
RV630
RV670
RV670
GPU
390M
666M
666M
Memory Size
256 MB
256 MB
512 MB
Memory Bus Width
128 bit
256 bit
256 bit
Core Clock
800 MHz
670 MHz
777 MHz
Memory Clock
1100 MHz
828 MHz
1126 MHz
Price
$105
$179
$249
Specification
666 million transistors on 55nm fabrication process
PCI Express x16 GPU bus interface
256-bit GDDR3/GDDR4 memory interface
Ring Bus Memory Controller
Fully distributed design with 512-bit internal ring bus for memory reads and writes
Microsoft® DirectX® 10 1 support
Shader Model 4.1
32-bit floating point texture filtering
Indexed cube map arrays
Independent blend modes per render target
Pixel coverage sample masking
Read/write multi-sample surfaces with shaders
Gather4 texture fetching
Unified Superscalar Shader Architecture
320 stream processing units
Dynamic load balancing and resource allocation for vertex, geometry, and pixel shaders
Common instruction set and texture unit access supported for all types of shaders
Dedicated branch execution units and texture address processors
128-bit floating point precision for all operations
