Motorola Corporation

Meltin Bell

L Rodney Goke

Michael G Gallup

William C Anderson

Michael F Wiles

Erik L Welty

Jack R Davis

Chuan-Chang Hung

Larry Yiucham Kan

A data processing system (55) and method thereof includes one or more data processors (10). Data processor (10) is capable of performing both vector operations and scalar operations. Using a single microsequencer (22), data processor (10) is capable of executing both vector instructions and scalar instructions. Data processor (10) also has a memory circuit (14) capable of storing both vector operands and scalar operands.

Method and apparatus for performing a vector skip instruction in a data

A method and apparatus for moving data in a parallel processing system (3). In one embodiment, a single instruction accesses one significant bit of information from each element in processing element array (80) and combines these bits into one designated element in global register file (50). The ordering of bits in vectors of global register file (50) associates each bit with an element of processing element array (80). Another single instruction distinguishes significant bit information from a particular vector in global register file (50) and transfers that information to an associated bit in storage circuits associated with each element in processing element array (80).

Method and apparatus for moving data in a parallel processor

The present invention, which is implemented in a data processing system, includes a method for performing a fuzzy logic operation, wherein the steps are: (1) performing a rule evaluation process on a plurality of fuzzy input values, wherein one or more of the plurality of fuzzy input values has a lower limit (zero) value, and wherein the rule evaluation process (i) implements a rulebase including a plurality of rules associating a plurality of fuzzy inputs with a plurality of fuzzy outputs, (ii) results in one or more fuzzy output values, and (iii) further includes processing the one or more of the fuzzy input values having the lower limit value with respect to the rulebase, and (2) processing the fuzzy input values having a non-lower limit value with respect to the rulebase, the processing of the fuzzy input values having the non-lower limit value performed separately from the processing of the one or more of the fuzzy input values having the lower limit value. Sorting of the non-lower limit values may be performed. Sorting may be performed using an insertion sort method wherein elements of an unsorted list stored within a vector register are compared to a growing vector of sorted values, and each element from the unsorted list is then inserted into the sorted list. Alternatively, sorting may be performed by simultaneously comparing all of the elements of an unsorted list and then transferring the largest value to a second list, wherein this process is repeated for each element in the unsorted list.

Circuit and method for processing lower limit value fuzzy inputs during

A method for complex data movement in a multi-processor data processing system. In one embodiment, the multi-processor data processing system (10) includes an array (12) of data processors (50-65), a plurality of edge interface circuits (14,16), and a bus interface controller (22). In an alternate embodiment, multiprocessor data processing system (210) includes an array (212) of data processors (250-258), a plurality of edge interface circuits (214-217), and a bus interface controller (222). The data processing systems (10,210) are capable of performing complex data movement patterns between the processors (50-65,250-258) and the corresponding edge interface circuits (14, 16, 214-217), such as a transpose pattern, a ping-pong pattern, and a checkerboard pattern.

Growing community of inventors

Austin, TX, United States of America

Meltin Bell