Data processor and data processing system

A data processor comprising:

• a floating point unit executable of inner product operation of 4.times.4, for which components are providing in floating point numbers by one floating point command;
• said floating point unit including an arithmetic block and a register file supplying objective data for arithmetic operation to said arithmetic block and being supplied a resultant data of arithmetic operation calculated by said arithmetic block;
• and
• said arithmetic block including four multipliers performing multiplication of mantissa parts of floating point numbers, aligners for receiving outputs of respective of said multipliers and performing alignment shift, an exponent processing portion for generating number of alignment shift of said aligners and exponents before normalization on the basis of exponent parts of said floating point numbers, a four input adder adding outputs of said aligners, and a normalizer for normalizing an output of said four input adder and said exponents before normalization.

A data processor comprising:

• a floating point number unit executable of calculation of a product of a 4.times.4 transformation matrix, components of which are given as floating point numbers respectively, and a vector having length 4 components respectively given as floating point numbers, by sequentially executing 4.times.4 inner product operation for four times, by one floating point command;
• said floating point unit including an arithmetic block and a register file supplying objective data for arithmetic operation to said arithmetic block and being supplied the resultant data of said arithmetic operation calculated by said arithmetic block;
• said arithmetic block including four multipliers performing multiplication of mantissa parts of floating point numbers, aligners receiving outputs from said multipliers for performing alignment shift, an exponent processing portion for generating number of alignment shift of said aligners and exponents before normalization on the basis of exponent parts of said floating point numbers, a four input adder adding the outputs of said aligners, and a normalizer for normalizing the output of said four input adder and said exponents before normalization;
• and
• said register file including two register banks respectively including sixteen registers, said transformation matrix being assigned to one of said bank, and said vector is assigned to the other bank.

A three-dimensional computer graphics processing system comprising:

• a data processor including an arithmetic portion incorporated in a floating point unit, including a plurality of multipliers supplied with a mantissa part of floating point number from respectively different data input signal line group and performing mutual multiplication of supplied mantissa parts, an aligner receiving outputs of respective multipliers and performing alignment shift, an exponent processing portion for generating number of alignment shift of said aligner and an exponent before normalization on the basis of generation an exponent part of said floating point number, a multi-input adder for adding the outputs of said aligner, and a normalizer for normalizing the output of said multi-input adder and said exponent before normalization;
• a rendering co-processor connected to said data processor;
• and
• said rendering co-processor performing a rendering processing of the three-dimensional computer graphics processing, and said data processor performing at least one of inner product operation and vector transforming operation for geometry processing of the three-dimensional computer graphics processing.