RSX
Hardware
RSX - Reality Synthesizer
The RSX is a graphical processor unit (GPU) based off of the nVidia 7800GTX graphics processor, and is a G70/G71 hybrid with some modifications. The RSX has separate vertex and pixel shader pipelines.
The following is a small sample of serial numbers of the RSX by model number.
| PS3 Model | RSX Serial | Die Tech | Die Size |
|---|---|---|---|
| CECHA | CXD2971GB | 90nm | 258mm2 |
| CECHC | CXD2971DGB | 90nm | 258mm2 |
| CECHG | CXD2971DGB | 90nm | 258mm2 |
| CECHH | CXD2971AGB | 90nm | 258mm2 |
| CECHK | CXD2982GB | 90nm | 258mm2 |
| CECHL | CXD2991GB | 65nm | ?mm2 |
| CECH-20xx | CXD2991EGB | 65nm | ?mm2 |
| CECH-21xx | CXD5300AGB | 40nm | ?mm2 |
| CECH-25xx | CXD5300A1GB | 40nm | ?mm2 |
The following are relevant facts about the RSX...
Little Endian
8 vertex shaders at 500Mhz
28 pixel shaders (4 redundant, 24 active) at 550Mhz
28 texture units (4 redundant, 24 active)
8 Raster Operations Pipeline units (ROPs)
Includes 256MB GDDR3 650Mhz clocked graphics memory
Earlier PS3 Models: Samsung K4J52324QC-SC14 rated at 700Mhz
Later PS3 Models: Qimonda HYB18H512322AF-14
GDDR3 Memory interface bus width: 128bit
Rambus XDR Memory interface bus width: 56bit out of 64bit (serial)
More features are revealed in the following chart delineating the differences between the RSX and the nVidia 7800 GTX.
| Difference | RSX | nVidia 7800GTX |
|---|---|---|
| GDDR3 Memory bus | 128bit | 256bit |
| ROPs | 8 | 16 |
| Post Transform and Lighting Cache | 63 max vertices | 45 max vertices |
| Total Texture Cache Per Quad of Pixel Pipes (L1 and L2) | 96kB | 48kB |
| CPU interface | FlexIO | PCI-Express 16x |
| Technology | 40nm/65nm/90nm | 110nm |
Other RSX features/differences include:
More shader instructions
Extra texture lookup logic (helps RSX transport data from XDR)
Fast vector normalize
Note that the cache (Post Transform and Lighting Vertext Cache) is located between the vector shader and the triangle setup.
A sample flow of data inside the RSX would see them first processed by 8 vertex shaders. The output are then sent to the 24 active pixel shaders, which can involve the 24 active texture units. Finally, the data is passed to the 8 Raster Operation Pipeline units (ROPs), and on out to the GDDR3. Note that the pixel shaders are grouped into groups of four (called Quads). There are 7 Quads, with 1 redundant, leaving 6 Quads active, which provides us with the 24 active pixel shaders listed above (6 times 4 equals 24). Since each Quad has 96kB of L1 and L2 cache, the total RSX texture cache is 576kB. General RSX features include 2x and 4x hardware anti-aliasing, and support for Shader Model 3.0.
Although the RSX has 256MB of GDDR3 RAM, not all of it is useable. The last 4MB is reserved for keeping track of the RSX internal state and issued commands. The 4MB of GPU Data contains RAMIN, RAMHT, RAMFC, DMA Objects, Graphic Objects, and the Graphic Context. The following is a breakdown of the address within 256MB of the RSX.
| Address Range | Size | Comment |
|---|---|---|
| 0000000-FBFFFFF | 252 MB | Framebuffer |
| FC00000-FFFFFFF | 4 MB | GPU Data |
| FF80000-FFFFFFF | 512KB | RAMIN: Instance Memory |
| FF90000-FF93FFF | 16KB | RAMHT: Hash Table |
| FFA0000-FFA0FFF | 4KB/s | RAMFC: FIFO Context |
| FFC0000-FFCFFFF | 64KB | DMA Objects |
| FFD0000-FFDFFFF | 64KB | Graphic Objects |
| FFE0000-FFFFFFF | 128KB | GRAPH: Graphic Context |
Speed, Bandwidth, and Latency
Because of the aforementioned layout of the communication path between the different chips, and the latency and bandwidth differences between the various components, there are different access speeds depending on the direction of the access in relation to the source and destination. The following is a chart showing the speed of reads and writes to the GDDR3 and XDR memory from the viewpoint of the Cell and RSX. Note that these are measured speeds (rather than calculated speeds) and they should be worse if RSX and GDDR3 access are involved because these figures were measured when the RSX was clocked at 550Mhz and the GDDR3 memory was clocked at 700Mhz. The shipped PS3 has the RSX clocked in at 500Mhz (front and back end, although the pixel shaders run separately inside at 550Mhz). In addition, the GDDR3 memory was also clocked lower at 650Mhz.
| Processor | 256MB XDR | 256MB GDDR3 |
|---|---|---|
| Cell Read | 16.8GB/s | 16MB/s |
| Cell Write | 24.9GB/s | 4GB/s |
| RSX Read | 15.5GB/s | 22.4GB/s |
| RSX Write | 10.6GB/s | 22.4GB/s |
Because of the VERY slow Cell Read speed from the 256MB GDDR3 memory, it is more efficient for the Cell to work in XDR and then have the RSX pull data from XDR and write to GDDR3 for output to the HDMI display. This is why extra texture lookup instructions were included in the RSX to allow loading data from XDR memory (as opposed to the local GDDR3 memory).
RSX Libraries
The RSX is dedicated to 3D graphics, and developers are able to use different API libraries to access its features. The easiest way is to use high level PSGL, which is basicially OpenGL|ES with programmable pipeline added in. At a lower level developers can use LibGCM, which is an API that talks to the RSX at a lower level. PSGL is actually implemented on top of LibGCM. For the advanced programmer, you can program the RSX by sending commands to it directly using C or assembly. This can be done by setting up commands (via FIFO Context) and DMA Objects and issuing them to the RSX via DMA calls.
Source: http://www.edepot.com/playstation3.html#PS3_RSX_GPU