利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

15 条记录 (耗时 0.052 秒)

    Computer implemented method for transferring packed data between register files and memory
    8.
    发明授权
    Computer implemented method for transferring packed data between register files and memory 失效
    用于在寄存器文件和存储器之间传送打包数据的计算机实现的方法

    公开(公告)号：US5935240A

    公开(公告)日：1999-08-10

    申请号：US573238

    申请日：1995-12-15

    申请人： Larry M. Mennemeier ,  Alexander D. Peleg ,  Andrew F. Glew ,  Carole Dulong ,  Eiichi Kowashi ,  Millind Mittal ,  Wolf Witt ,  Benny Eitan

    发明人： Larry M. Mennemeier ,  Alexander D. Peleg ,  Andrew F. Glew ,  Carole Dulong ,  Eiichi Kowashi ,  Millind Mittal ,  Wolf Witt ,  Benny Eitan

    IPC分类号： G06F13/16 ,  G06F9/305

    CPC分类号： G06F13/16

    摘要： A method for transferring packed data including the steps of first receiving an instruction from a set of instructions for transferring packed data between an extended register file and either an integer register file or a memory. In one embodiment, the extended register file includes eight registers, with each of the extended register storing up to sixty-four data bits. The integer register file also includes eight registers. The instruction includes an opcode that specifies a direction of the transfer with respect to the extended register file. The instructions are encoded in an instruction format having up to three bits addressing a destination operand and up to three bits addressing a source operand. The instruction is then translated to determine a direction of the transfer, a size of said packed data to be transferred, the address of the destination operand, and the address of the source operand. The instruction decoded by a decoder unit previously designed to decode the instruction format used to encode the set of instructions. In response to receiving the instruction, the packed data is transferred between the extended register file and either the integer register file or the memory, pursuant to the specifications of the translated instruction.

    摘要翻译： 一种用于传送打包数据的方法，包括以下步骤：首先从扩展寄存器文件和整数寄存器文件或存储器之间传送打包数据的一组指令接收指令。 在一个实施例中，扩展寄存器文件包括八个寄存器，每个扩展寄存器最多存储64个数据位。 整数寄存器文件还包括八个寄存器。 该指令包括指定关于扩展寄存器文件的传送方向的操作码。 指令以指令格式进行编码，最多三位寻址目标操作数，最多三位寻址源操作数。 然后转换该指令以确定传送的方向，要传送的所述打包数据的大小，目的地操作数的地址和源操作数的地址。 由解码器单元解码的指令，其先前设计为解码用于编码指令集的指令格式。 响应于接收到该指令，根据翻译指令的规范，打包数据在扩展寄存器文件和整数寄存器文件或存储器之间传送。