Special Winter Seminar Series - Quantum Computing and Information
Implementing Variational Algorithms on Near-term Quantum Processors
Alex McCaskey
Oak Ridge National Lab
Programmable quantum processors are available from various hardware vendors over the cloud with corresponding open source software frameworks enabling the programming, compilation, and execution of small-scale quantum circuit experiments at the assembly level. This has opened up new research avenues that seek to explore the potential capabilities that these processors may provide in the near-term for scientific computing, and specifically what algorithms may prove useful for advancing the state-of-the-art with regards to scientific simulation tasks. In this talk, I will detail useful hybrid variational algorithms that leverage classical optimization in tandem with quantum co-processor subroutine execution, how we have leveraged these algorithms in the presence of noise on remotely hosted quantum processors, and novel work enabling the implementation of these algorithms at levels familiar to domain computational scientists. I will demonstrate experimental results for nuclear physics, quantum chemistry, and machine learning, and detail methods and strategies employed for coarse-grained error mitigation via post-processing of noisy execution results. Ultimately, this talk should provide a high-level background on near-term quantum computing, variational algorithm design and implementation, and its overall applicability to certain domain scientific applications.
All interested persons are invited to attend remotely—email physics@nd.edu for information.