Research On Quantum Compiler Optimization Method Based On Domain Specific Language

In recent years,with the rapid development of the information industry and the continuous popularization and deepening of Internet technology,people are gradually entering the era of big data,which will intensify the global dependence on computing power.On the one hand,the interaction between classical computing and quantum computing is a hot topic in the quantum field.The use of domainspecific languages to describe heterogeneous quantum-classical computing algorithms can accelerate the research on quantum computing by experts and software developers in the quantum field.process.On the other hand,directly converting quantum basic logic instructions into control pulses will result in lowefficiency and high-latency programs.This thesis optimizes the quantum compilation method,which can reduce the pulse duration and generate highly optimized control pulses.main tasks as follows:1.Aiming at the difficulty of interaction between classical computing and quantum computing,design a quantum domain special language QCDSL(Quantum Compile Domain Specific Language).QCDSL is used to describe heterogeneous quantum-classical computing algorithms and algorithms in the process of quantum compilation and optimization.This paper analyzes and implements the design principles,module system,type system and configuration files of QCDSL,and conducts specific research on the key technologies of the proposed QCDSL quantum programming framework,including quantum program cycle,running system,main program and quantum compilation.At the same time,in QCDSL,a mechanism to support interaction with quantum algorithms is introduced.This mechanism uses internally defined operations and generates a quantum assembly instruction set.The experimental results show that the quantum programming framework of QCDSL realizes the rapid interaction between classical computing and quantum computing,laying a foundation for the co-optimization of quantum and classical computing.2.Aiming at the problem of low efficiency and high latency programs caused by traditional quantum compilation,this thesis proposes a multi-instruction aggregation quantum compilation optimization method.First,analyze the Max Cut problem of the QAOA circuit.After that,QCDSL is used in the front end of the quantum compilation to realize the exchangeable logic scheduling algorithm,and the exchangeability test of the quantum circuit is carried out.Subsequently,experimental research on the instruction aggregation method of the back-end of quantum compilation is carried out.Finally,several important classical quantum hybrid algorithms are selected as benchmarks,and the optimization method proposed in this thesis is compared with the unoptimized compilation method.The experimental results show that exchangeability detection and instruction aggregation can reduce the normalization delay of the circuit,and the delay effect is proportional to the exchangeability of the circuit,and the aggregation instruction compilation can greatly improve the circuit with low spatial locality.For the maximum cutting problem of QAOA,the quantum compilation method of multiinstruction aggregation can generate highly optimized control pulses,reducing the pulse duration by about 2.21 times.