arrow-up icon
Fixstars Amplify user

Juri Minami

Department of Physics/Department of Applied Physics, School of Advanced Science and Engineering, Waseda University
Yorita's research group (Particle physics)

I'm a member of the Yorita's research group. Our group participates in the ATLAS (particle accelerator) experiment at the European Organization for Nuclear Research (CERN), one of the world's largest particle physics laboratories located in Switzerland. We conduct daily research to unravel the details of the Higgs boson, whose existence was confirmed in 2012, and to search for unknown particles.

In our research, we investigate the characteristics of particles by the track reconstruction after high-energy proton-proton collisions, and the amount of data we obtain at ATLAS is enormous. This data is expected to increase further with the planned updates to the accelerator and detectors. The more data we have, the more computational time and resources are needed to calculate the track reconstruction. To deal with these challenges, we are considering the use of a quantum computer or an Ising machine, and chose to use Fixstars Amplify, because their web site has a very good introduction and documentation.

In our recent research, we considered the track reconstruction as a QUBO-style combinatorial optimization problem and solved it using the Fixstars Amplify AE. We confirmed that we could obtain a solution with a certain level of accuracy in a very short annealing time. There are still some issues to be considered, such as how to improve the accuracy by adjusting the parameters of the function used to create the QUBO matrix, and how to obtain a solution with the same accuracy even when the number of data increases.

Even though I started using Python in my master's program, I was able to easily master Fixstars Amplify since their website has many useful technical contents such as demo & tutorials, sample codes, and API documents. I think one of the benefits of Fixstars Amplify is that I can switch the machines very easily. It should help me to run a benchmark among different machines in our future research and find the best annealing machine to solve my research problem. I hope that Fixstars Amplify will keep improving the accessibilities to the other machines and solvers through the unified middleware, Fixstars Amplify SDK.

*All information in this article is based on information available at the time of the interview (March 2023)


参考資料
論文名

Image

Results of the track reconstruction obtained with Amplify.