Presentation Information

[7p-P07-2]Development and Benchmarking of a Computational Simulation System for Unconventional Superconductivity Coded in Julia

〇Naotaka Tanaka1, Kensei Ushio1, Daiki Nakaoka1, Yuto Hoshi1, Hirofumi Sakakibara1,2, Masataka Kakoi3 (1.Faculty of Eng., Tottori Univ., 2.AMES., Tottori Univ., 3.Dept. of Phys., Osaka Univ.)

Keywords:

Superconductivity,Electron Correlation,First-Principles Calculations

This research focuses on developing an analysis code that enables high-precision evaluation of physical properties for the exploration of novel superconducting materials. Electronic state analysis based on first-principles band calculations is essential for elucidating material properties and predicting new superconductors. By combining this with the Fluctuation-Exchange (FLEX) approximation—a simulation technique grounded in many-body electron theory—it becomes possible to explore candidate materials without referring experimental data. In this framework, the superconducting transition temperature can be predicted by solving the linearized Eliashberg equation and evaluating its eigenvalues. The code is being developed in the fast and highly extensible Julia language. In the presentation, we will outline the structure of the code and provide performance benchmarks.