量子程序中一种改进的延时估计算法研究

Research on improved latency estimation algorithm in quantum program

针对现有的量子算法处理时间长、成本过于昂贵这一不足,提出一种延时快速估计算法LEQA,以评估映射到瓦片量子结构的量子算法的延时。该算法使用程序化方法,通过计算量子周围相邻量子的数量来计算算法的延时。仿真结果表明,对中等规模的电路,LEQA的速度要比执行细致的布局、调度以及将量子运算依赖图中的量子指令和量子位通过路由传输到量子结构中的现代量子映射器速度快出两个数量级。此外,LEQA还能快速得出电路延时的估计,且精度很高,平均误差只有2.11%。

Aiming at the disadvantages of the processing time was too long and the cost was too high in existing quantum algorithms, this paper proposed a fast delay estimation algorithm named LEQA to assess the latency of the quantum algorithms, which were mapped to tile quantum structure. The algorithm used a procedural method. It counted the latency of the algorithm by calculating the number of adjacent quantum. Simulation results show that, for medium-sized circuits, the speed of LEQA was two orders of magnitude faster than that of modern quantum mapper, which performed detailed layout, scheduling and transmits quantum instructions, which was existing in the dependency graph of quantum computing, and qubit to the quantum structure by routing. In addition, LEQA can produce quick estimates of the circuit latency with sufficient accuracy, and the average error is only 2.11%.