PARALLELIZATION OF SCALAR OPERATIONS BY VECTOR PROCESSORS USING DATA-INDEXED ACCUMULATORS IN VECTOR REGISTER FILES, AND RELATED CIRCUITS, METHODS, AND COMPUTER-READABLE MEDIA

摘要

Parallelization of scalar operations by vector processors using data-indexed accumulators in vector register files, related circuits, methods, and computer-readable media are disclosed. In one aspect, a vector processor comprises a vector register file providing a plurality of write ports and a plurality of vector registers each providing a plurality of accumulators. The vector processor receives an input data vector. For each of the plurality of write ports, the vector processor executes vector operation(s) for accessing an input data value of the input data vector, and determining, based on the input data value, a register index for a vector register among the plurality of vector registers, and an accumulator index for an accumulator among the plurality of accumulators of the vector register. Based on the register index, a register value is retrieved from the register index, and a scalar operation is performed based on the register value and the accumulator index.

主权项

1. A vector processor, comprising:
a vector register file comprising a plurality of vector registers each configured to provide a plurality of accumulators; the vector register file configured to provide a plurality of write ports; and the vector processor configured to:
receive an input data vector; andexecute one or more vector operations for each write port of the plurality of write ports to:
access an input data value of the input data vector;determine, based on the input data value, a register index indicative of a vector register among the plurality of vector registers;determine, based on the input data value, an accumulator index indicative of an accumulator among the plurality of accumulators of the vector register; andbased on the register index and the accumulator index, perform a scalar operation on the vector register indicated by the register index.