A Streaming Multi-GPU Implementation of Image Simulation Algorithms for Scanning Transmission Electron Microscopy

TL;DR

Prismatic is a GPU-accelerated software package that significantly speeds up atomic resolution image simulations in STEM, enabling practical use in electron tomography with up to 1000x acceleration.

Contribution

This work introduces Prismatic, a parallelized GPU software that accelerates STEM image simulations using PRISM and multislice methods, facilitating real-time applications.

Findings

Achieves up to 1000x acceleration with PRISM

Achieves up to 30x acceleration with multislice

Enables faster atomic electron tomography simulations

Abstract

Simulation of atomic resolution image formation in scanning transmission electron microscopy can require significant computation times using traditional methods. A recently developed method, termed plane-wave reciprocal-space interpolated scattering matrix (PRISM), demonstrates potential for significant acceleration of such simulations with negligible loss of accuracy. Here we present a software package called Prismatic for parallelized simulation of image formation in scanning transmission electron microscopy (STEM) using both the PRISM and multislice methods. By distributing the workload between multiple CUDA-enabled GPUs and multicore processors, accelerations as high as 1000x for PRISM and 30x for multislice are achieved relative to traditional multislice implementations using a single 4-GPU machine. We demonstrate a potentially important application of Prismatic, using it to compute images for atomic electron tomography at sufficient speeds to include in the reconstruction pipeline. Prismatic is freely available both as an open-source CUDA/C++ package with a graphical user interface and as a Python package, PyPrismatic.