International Business Machines Corporation

Clair John Glossner, Iii

Mayan Moudgill

David Meltzer

Erdem Hokenek

Erik Richter Altman

A system and method for processing operations that use data vectors each comprising a plurality of data elements, in accordance with the present invention, includes a vector data file comprising a plurality of storage elements for storing data elements of the data vectors. A pointer array is coupled by a bus to the vector data file. The pointer array includes a plurality of entries wherein each entry identifies at least one storage element in the vector data file. The at least one storage element stores at least one data element of the data vectors, wherein for at least one particular entry in the pointer array, the at least one storage element identified by the particular entry has an arbitrary starting address in the vector data file.

Vector register file with arbitrary vector addressing

A system and method is provided for processing a first instruction set and a second instruction set in a single processor. The method includes storing a plurality of instructions of the second instruction set in a plurality of buffers proximate to a plurality of execution units, executing an instruction of the first instruction set in response to a first counter, and executing at least one instruction of the second instruction set in response to at least a second counter, wherein the second counter is invoked by a branch instruction of the first instruction set.

System and method including distributed instruction buffers for storing frequently executed instructions in predecoded form

A digital signal processor (DSP) includes dual SIMD units that are connected in cascade, and wherein results of a first SIMD stage of the cascade may be stored in a register file of a second SIMD stage in the cascade. Each SIMD stage contains its own resources for storing operands and intermediate results (e.g., its own register file), as well as for decoding the operations that may be executed in that stage. Within each stage, hardware resources are organized to operate in SIMD manner, so that independent SIMD operations can be executed simultaneously, one in each stage of the cascade. Intermediate operands and results flowing through the cascade are stored at the register files of the stages, and may be accessed from those register files. Data may also be brought from memory directly into the register files of the stages in the cascade.

Digital signal processor with cascaded SIMD organization

A volatile memory device, in accordance with the present invention, includes an array of memory cells with at least two dummy cells disposed within the memory array. A driver is included for writing a first state to one of the at least two dummy cells and for writing a second state to another one of the at least two dummy cells. A comparison circuit compares the first state and the second state to a threshold to determine if a refresh of the array of memory cells is needed.

Growing community of inventors

Carmel, NY, United States of America

Clair John Glossner, Iii