首页> 外文会议>2012 IEEE Asia Pacific Conference on Circuits and Systems. >Wirelength driven I/O buffer placement for flip-chip with timing-constrained
【24h】

Wirelength driven I/O buffer placement for flip-chip with timing-constrained

机译:用于时序限制的倒装芯片的线长驱动I / O缓冲区放置

获取原文
获取原文并翻译 | 示例

摘要

Flip-chip package provides the highest chip density because I/O buffers in it could be placed anywhere inside a chip. The assignment of bump pads, I/O buffers and I/O pins will affect the satisfaction of timing requirement inside die core. In this paper, we proposed an effective three-step hierarchical approach to satisfy the timing-constrained I/O buffer placement in an area-I/O flip-chip design, meanwhile, wirelength could be optimized. First of all, I/O buffers are inserted to the floorplan plane greedily, and then, the wirelength between I/O buffers and I/O pins are optimized by a fixed-outline floorplanning algorithm. Secondly, a network flow model is conducted, and a min-cost-max-flow algorithm is used to assign I/O pins, I/O buffers and bump pads. Finally, the timing constraints are translated to length constraints, the results that satisfy timing constraints are selected. The experimental results show that, under the given timing constraints, higher timing-constrained satisfaction ratio (TCSR) is obtained, and the reduction of total wirelength is 14% on average.
机译:倒装芯片封装可提供最高的芯片密度,因为其中的I / O缓冲区可以放置在芯片内部的任何位置。缓冲垫,I / O缓冲器和I / O引脚的分配将影响芯片核心内部对时序要求的满足。在本文中,我们提出了一种有效的三步分层方法来满足区域I / O倒装芯片设计中受时间限制的I / O缓冲区的放置,同时可以优化线长。首先,将I / O缓冲区贪婪地插入到平面图平面中,然后,通过固定轮廓布局规划算法来优化I / O缓冲区和I / O引脚之间的线长。其次,进行网络流模型,并使用最小成本最大流算法分配I / O引脚,I / O缓冲区和缓冲垫。最后,将时序约束转换为长度约束,选择满足时序约束的结果。实验结果表明,在给定的时序约束下,可以获得较高的时序约束满意率(TCSR),平均总导线长度减少14%。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号