Calculation Of Phonon Scattering Rate And Prediction Of Lattice Thermal Conductivity With GPU

Lattice thermal conductivity(LTC) is a key parameter for many technological applications.Based on the Peierls–Boltzmann transport equation(PBTE),many unique phonon transport properties of various materials were revealed.Accurate calculation of LTC with PBTE,however,is a time-consuming task,especially for compounds with a complex crystal structure.At high temperatures,the high-order phonon scattering process will also have a significant impact on the thermal conductivity of the lattice,and the amount of calculation will increase by orders of magnitude.Graphical processing units(GPUs)have been extensively used to accelerate scientific simulations,which greatly improve computational efficiency.Researchers can use a single desktop workstation equipped with a GPU to complete computing tasks that used to require supercomputers.Due to the fundamental differences between the architecture of GPU and CPU,GPUs are especially suited for performing parallel computing but require specified algorithm design.In this paper,we propose a new algorithm optimized based on GPU architecture,where the phonon scattering rate under a given phonon mode will be accumulated in a single thread to avoid warp divergence.A new open-source code,GPU＿PBTE,is released based on the proposed algorithm.Furthermore,we used this tool to calculate the lattice thermal conductivity of silicon,silicon carbide,single-layer graphene and filled skutterudite,and analyzed the influence of phonon scattering processes on its lattice thermal conductivity.We compare the calculation results with the experimental data and find that accurate and reliable LTC can be obtained with the software.GPU＿PBTE executed on NVIDIA Tesla V100 can broadly improve the double-precision performance,making it two to three orders of magnitude faster than the CPU version performed on Intel Xeon CPU Gold6248 @2.5 GHz.This work provides strong support for exploring the lattice thermal conductivity of complex crystals and for designing new thermal functional materials.It also provides ideas for accelerating calculations with other novel hardware that may come out in the future.