发明名称 DATA PROCESSING APPARATUS AND METHOD FOR PERFORMING VECTOR SCAN OPERATION
摘要 A vector scan operation is performed to generate M data elements of a result vector, where each result data element corresponds to a combination of an additional data element with at least some of the data elements of a source vector operand V. The vector scan operation is performed using a plurality of steps, each step comprising one or more combination operations for combining data elements. At least one of the steps includes two or more combination operations performed in parallel. At least two of the steps comprise a combination operation for combining a data element with the additional data element S. This approach enables the vector scan operation to be performed in fewer steps in the case where fewer than M data elements are active, so that the vector scan operation can be performed more quickly.
申请公布号 US2015254076(A1) 申请公布日期 2015.09.10
申请号 US201514601431 申请日期 2015.01.21
申请人 ARM LIMITED 发明人 BOETTCHER Matthias;GABRIELLI Giacomo;EYOLE-MONONO Mbou
分类号 G06F9/30 主分类号 G06F9/30
代理机构 代理人
主权项 1. A data processing apparatus comprising: a vector register store configured to store vector operands comprising a plurality of data elements; processing circuitry configured to process vector operands from the vector register store; and control circuitry configured to control the processing circuitry to perform a vector scan operation on M data elements of a source vector operand V[0] to V[M−1] and at least one additional data element S, to generate M data elements of a result vector operand R[0] to R[M−1], where for N≦M and 0≦i<N, data element R[i] of the result vector operand has a value corresponding to a combination of the at least one additional data element S and at least some of data elements V[0] to V[i] of the source vector operand; wherein the control circuitry is configured to control the processing circuitry to perform the vector scan operation in a plurality of steps, each step for generating a second vector from a first vector, where the first vector for a first step comprises data elements of the source vector operand, and the first vector for other steps comprises the second vector of the preceding step, each step comprising at least one combination operation for combining a data element of the first vector with the at least one additional data element S or another data element of the first vector to generate a data element of the second vector; at least one of said plurality of steps comprises a plurality of combination operations performed in parallel; and at least two of said plurality of steps comprise a combination operation for combining a data element of the first vector with the at least one additional data element S.
地址 Cambridge GB