First SAP Collaboration a Quantum Success

By Herbert Morgan, NCSA

As the first Strategic Application Program (SAP) project draws to a close, the collaborators are pleased with their results. The SAP project "High Performance Algorithms for Scalable Spin-Qubit Circuits with Quantum Dots" materialized almost six months ago when Jean-Pierre Leburton, the project's principal investigator, met with Nahil Sobh, the SAP coordinator.

The project seeks to understand how to build new kinds of high-performance devices based on the manipulation of the electron spin that defines a quantum bit or qubit in quantum dots for quantum information processing. This project was considered ideal for SAP because it involved a small research group that had large computational needs.

Working with Leburton was Dmitriy Melnikov, a postdoc research associate in the Department of Electrical and Computer Engineering. Working with Sobh was John Fettig, a graduate research assistant at NCSA. As far as the project was concerned, Melnikov and Fettig were the primary collaborators. Fettig was asked to do two things: optimize the code and get it to run in parallel.

"The code that Dmitriy was working with," says Fettig, "had several iterative cells in it. So I took those iterative cells and replaced them with a library that is much faster. The library is also parallel so the code now will run on probably up to eight processors. It hits a limit on scalability for the problem size that was tested."

According to Melnikov, "What we are doing is simulating basic device operations for quantum computing in which modeling of spin states is tedious. Therefore, for many spin operations, program speed is very essential."

Initially, Fettig replaced the entire eigenvalue solver but found that it didn't work. He also expressed disappointment that he couldn't get the ARPACK (Arnoldi Package) or SLEPc (Scalable Library for Eigenvalue Problem Computations) to work instead of just replacing the conjugate gradient method. But to his delight, the results exceeded his expectations: "I never dreamed it would have sped up the code as much as it did."

Melnikov concurs, "What was taking 24 hours on a single processor now takes 2 hours. [Fettig's contribution] was very helpful, and on top of that we have parallel code now."

Because of the optimization gains and the parallel capabilities, Melnikov can now simulate two or three quantum dots, instead of just one. He explained that in terms of size, there are two problems: one is real and the other is complex. The real is manageable even for two quantum dots. But the complex problem becomes unmanageable with just two quantum dots.

About the first SAP project, Sobh says, "I think it went very well. We would like to continue and explore the complex solvers as well as the data and visualization aspects."

The collaborators met weekly and communicated through email. They believe that it helped to meet in person, especially in the initial stages of the project.