[15p-P07-1~60] 10 Spintronics and Magnetics (Poster)

[15p-P07-25] GPU implementation of quasiparticle self-consistent GW calculations: Application to the modulated martensite phase of magnetic shape memory alloy Ni₂MnGa

〇Masao Obata¹, Takumi Katsuta¹, Takao Kotani², Tatsuki Oda¹ (1.Kanazawa Univ., 2.Tottori Univ.)

Keywords:

Ferromagnetic shape memory alloy,GW calculation,GPU implementation

Ni-Mn-Ga alloys possess ferromagnetic and shape memory properties, making them promising for applications in magnetically induced actuators. At low temperatures, their crystal structure forms a modulated martensite structure that includes nano-twin boundaries. By utilizing GPU, we accelerated the calculation of self-consistent quasiparticle GW, which is a highly reliable first-principles electronic structure calculation. With the aid of GPU-accelerated QSGW calculation, we analyzed the electronic structure for the modulated martensite phase of Ni₂MnGa. The electronic density of states near the Fermi level forms a valley structure, which is considered one of the factors in stabilizing the modulated structures.