Sapphire X1900XT-X 512mb
Author: Matthew Kemp (kempez815)
Hardware Acquired: SCAN Computers
Meet the Chip – R580
ATI’s X1900XT-X is based upon their newest 90nm chip the R580. with all-new features and specifications this chip looks like it could very well be something special.
|DirectX Shader||SM 3.0|
|Vertex Shader||VS3.0 (no texld)
5D FP32, co-issue MADD, branch
single cycle trig functions
4D FP32, dual-issue ADD+MADD, branch
single cycle trig function
|Texture Processing||16 FP32 address units, 4 samplers
Bilinear filter for integer sample
|Memory||256-bit, 512-bit ring bus
|Output||2 x Dual Link DVI Ports|
All of these numbers are a bit confusing so I'll give a brief summary of which parts are new in the R580.
ALU's and R580
The R580 differs the most from its pre-dessors by its ALU (Arithmetic Logic Unit). Numbers get transferred from memory into the ALU for calculation, and the results are sent back into memory. The R580 has 48 ALU's.
Basically this architecture allows the R580 to send theoretically three times the fragment data than the R530 (X1800XT). Thats three times the data. This even tops the Xenos from the XBox 360 and means that this makes the X1900XT-X the fastest GPU around right now.
The X1900XT-X shares the 512-bit internal, 256-bit external ringbus memory architecture that the X1800XT and family use. Four ring are for the DRAM devices, which connect to node in pairs. The last is for general I/O to externals like the PCI Express bus and ATI HyperMemory, allowing the memory controller to address those resources. This controller allows the X1900XT-X to be flexible in the way it talks to its memory and the hardware surrounding it and to be able to adapt easily to different clock speeds, with reduced latency.
Hexus explain this perfectly:
"The raster output hardware is the final piece of the 3D puzzle for R580. Each ROP, of which there's sixteen, can do two Z-writes per cycle (sustainable with MSAA on, to boot) or combined Z and colour. They also have the same programmable pixel sub-sampling for multisample antialiasing (MSAA, up to 6x) as R4-series GPUs, but can also pair that with sub-sampling of alpha textures to effectively antialias those surfaces, too.
Additionally, and to create one of the defining features of the R5 generation of hardware, the ROP can also multisample from 'HDR' surface formats. 64-bit per pixel surfaces - integer (FX16) or floating-point (FP16) - can be sampled from, along with 32-bit FX10 (10bpc RGB, 2-bit alpha), L16 (16-bit integer, luminance only) and combinations of those depending on what the developer needs.
The 32-bit FX10 format is the full-speed format for 'HDR' MSAA, and the surfaces can be sampled while compressed. It's a feature not present on any competing hardware for NVIDIA and gives ATI a large competitive advantage overall in the image quality stakes. Let's talk about image quality quickly, before moving on to the physical implementation of the first R580-based SKUs."
So basically the doubled up 32bit processing can rendor HDR and other DX 9.0c features faster and better than anything out there.
And it shows....