流水的浮点倒数近似值运算部件的设计与实现
作者:
作者单位:

(上海高性能集成电路设计中心, 上海 201204)

作者简介:

何军(1980—),男,湖北汉川人,高级工程师,博士,E-mail:joyhejun@126.com

通讯作者:

中图分类号:

TP332.2

基金项目:

国家核高基重大专项资助项目(2018ZX01029101)


Design and implementation of pipelined floating-point reciprocal approximation operation unit
Author:
Affiliation:

(Shanghai High Performance Integrated Circuit Design Center, Shanghai 201204, China)

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    在部分低精度浮点运算应用中,需要流水的浮点倒数近似值运算。本文基于SRT-4算法设计并实现了一种流水的浮点倒数近似值运算部件。该部件采用6级流水线结构,运算结果精度至少为8位有效尾数。为了支持对非规格化浮点数的硬件处理,还设计并实现了改进版,有利于进一步提高浮点倒数近似值运算的性能。改进版采用8级流水线结构,新增了源操作数预规格化和结果后规格化功能模块,可以实现对非规格化浮点数的硬件处理。经过逻辑综合评估,改进版的硬件开销是面积在合理范围内增加19.23%,且对时序没有明显影响,可以满足预期的1.6 GHz频率设计目标。

    Abstract:

    In some low precision applications, pipelined floating-point reciprocal operation is required actually. Based on SRT-4 algorithm, a pipelined floating-point reciprocal operation unit was designed and implemented, which is constructed as a 6-stage pipeline unit, resulting in an 8-bit valid fractions. In order to support hardware process of denormal numbers, the unit was improved to get higher performance, which is constructed as a 8-stage pipeline unit, adding source operand pre-normalization and result post-normalization function components and supporting hardware process of denormal numbers. After logic synthesis, the area of the unit was increased by 19.23%, which is reasonable. The timing of the unit was not affected obviously and met the expected frequency goal of 1.6 GHz.

    参考文献
    相似文献
    引证文献
引用本文

何军,王丽.流水的浮点倒数近似值运算部件的设计与实现[J].国防科技大学学报,2020,42(2):41-46.
HE Jun, WANG Li. Design and implementation of pipelined floating-point reciprocal approximation operation unit[J]. Journal of National University of Defense Technology,2020,42(2):41-46.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2019-09-30
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2020-04-29
  • 出版日期: 2020-04-28
文章二维码