发明名称 Functional unit for vector leading zeroes, vector trailing zeroes, vector operand 1s count and vector parity calculation
摘要 A method of performing vector operations on a semiconductor chip is described. The method includes performing a first vector instruction with a vector functional unit implemented on the semiconductor chip and performing a second vector instruction with the vector functional unit. The first vector instruction is a vector multiply add instruction. The second vector instruction is a vector leading zeros count instruction.
申请公布号 US9092213(B2) 申请公布日期 2015.07.28
申请号 US201012890457 申请日期 2010.09.24
申请人 Intel Corporation 发明人 Wiedemeier Jeff;Samudrala Sridhar;Golliver Roger;Mahurin Eric W.
分类号 G06F7/38;G06F9/00;G06F9/44;G06F15/00;G06F9/30 主分类号 G06F7/38
代理机构 Nicholson De Vos Webster & Elliot LLP 代理人 Nicholson De Vos Webster & Elliot LLP
主权项 1. A vector functional unit implemented on a semiconductor to perform vector operations of dimension N, comprising: N functional units, each of the N functional units comprising: a multiplier logic circuit and an adder logic circuit to perform a multiply add operation on respective A, B and C operands responsive to a vector multiply add instruction, the A, B and C operands received on respective inputs of the functional unit; a leading ones count logic circuit downstream from an alignment shifter circuit that precedes said adder logic circuit, said leading ones count logic circuit to perform a leading ones count operation responsive to the vector multiply add instruction to adjust an exponent value for the vector multiply add instruction; and a leading zeros count logic circuit downstream from said alignment shifter circuit to perform a leading zeros count operation on an operand received on one of the respective A, B and C operand inputs, the leading zeros count operation being responsive to a vector leading zeros count instruction.
地址 Santa Clara CA US