Processor efficient transformation and lighting implementation for three dimensional graphics utilizing scaled conversion instructions
Efficient techniques for computation of texture coordinates using scaled conversion operations for a 3D graphics pipeline utilizing a scaled floating point to integer instruction and a scaled integer to floating point instruction to significantly reduce memory requirements. A parallel array VLIW digital signal processor is employed along with specialized scaled conversion instructions and communication operations between the processing elements, which are overlapped with computation to provide very high performance operation. Successive iterations of a loop of tightly packed VLIWs are used allowing the graphics pipeline hardware to be efficiently used.
1. An apparatus for executing a programmable floating point to integer scaled (FTOIS) instruction, the apparatus comprising:
a first input for receiving an FTOIS instruction, the FTOIS instruction including a target data type;
a second input for receiving a floating point number, the floating point number including fraction bits and exponent bits;
a floating point conversion module having a saturation and result formatting unit, the floating point conversion module calculating a fraction shift amount based on the exponent bits and shifting the fraction bits by the fraction shift amount plus an offset determined by the target data type, the floating point conversion module rounding the shifted fraction bits, the saturation and result formatting unit generating overflow and underflow results depending on the target data type and whether there is an overflow or underflow situation, the saturation and result formatting unit determining a scaled integer number by multiplexing the rounded bits with either overflow or underflow results, said scaled integer number comprising fewer bits than the input floating point number.
2. The apparatus of claim 1 wherein the floating point number comprises a texture coordinate.
3. The apparatus of claim 1 wherein the scaled integer number is stored in a data type format selected from a plurality of data types.
4. The apparatus of claim 3 wherein the selected data type format is a byte in the range of 0 to 255.
5. The apparatus of claim 3 wherein the scaled integer number is duplicated depending on the selected data type format.
6. The apparatus of claim 1 wherein the floating point number is a single precision number in the range of 0.0 to 1.0.