基于GPU的VLSI的DRC加速系统  

DRC Accelerated System of VLSI Based on GPU

作　　者：池凤彬[1] 潘日华 陈扉 赵冬晖[1] 

机构地区：[1]复旦大学软件学院,上海200433 [2]圣景微电子(上海)有限公司,上海200433

出　　处：《微电子学与计算机》2007年第4期171-173,共3页Microelectronics & Computer

摘　　要：在超大规模集成电路(VLSI)设计流程中,设计规则检查(DRC)是关键一环。多年来,设计人员为DRC设计了许多硬件加速的方法,但是都局限于成本等诸多原因而不能得到推广。因此提出了基于GPU平台的DRC方法,大幅提高了DRC效率。In VLSI design flow, design rule checking （DRC） is an important step. In the past, designers have proposed many methods which are based on hardware. However they are often limited by the cost or some other reasons. Because of DRC building on GPU, the high performance floating-point Single Instruction Multiple Data（SIMD）, the Vector Arithmetic we proposed and a/so the parallel of GPU and CPU, the efficiency is markedly improved.

关 键 词：VLSI DRC GPU SIMD 向量算法 

分 类 号：TP31[自动化与计算机技术—计算机软件与理论]