A processor interprets a stream of data as instructions to execute; it has one or more processing units that perform integer and floating-point arithmetic as well as more advanced computations.
处理器将数据流解释为要执行的指令，它拥有一个或多个处理单元，用于执行整数和浮点运算以及更高级的计算。
These workloads heavily use either the floating-point units or the memory bandwidth.
这些工作负荷会占用大量的浮点单元或内存带宽。
A physical processor is organized as different execution units at the hardware level, for example, fixed point and floating point operation units.
在硬件级别中，物理处理器是作为不同的执行单元进行组织的，例如定点和浮点操作单元。
For optimizing floating-point units, a design based on pipeline techniques is described in this paper.
运用流水线技术对单精度浮点乘法和加法运算单元进行了优化设计。
For optimizing floating-point units, a design based on pipeline techniques is described in this paper.
运用流水线技术对单精度浮点乘法和加法运算单元进行了优化设计。
应用推荐