Abstract:

In order to fast get the optimal placement result in the global scope, this paper presents a placement algorithm based on quantum model. Combined with the traditional simulated annealing algorithm, it can make Field Programmable Gate Array(FPGA) placement significantly faster. Test results show that the placement operation speed of this algorithm improves average 2 or more times than the VPR placement algorithm on average, while improving the timing performance by 2%. With the increasing scale of FPGA devices and the circuits, the algorithm can effectively improve the FPGA software operating efficiency.

Key words: Field Programmable Gate Array(FPGA), placement algorithm, quantum model, simulated annealing algorithm

摘要：

为能在全局范围内快速搜索到优化的布局结果，提出一种基于量子模型的布局算法，并结合传统模拟退火算法实现FPGA布局。测试结果表明，相比VPR布局算法，该算法的布局运行速度平均提高了2倍以上，时序性能提升了2%，且随着FPGA芯片和电路规模的不断增大，能有效提高FPGA的软件运行效率。

关键词: 现场可编程门阵列, 布局算法, 量子模型, 模拟退火算法

CLC Number: 

• TN47