GAPD: a GPU-accelerated atom-based polychromatic diffraction simulation code.

Area of Science:

• Materials Science
• Computational Physics
• Crystallography

Background:

• Accurate simulation of X-ray and electron diffraction is crucial for understanding atomic structures.
• Simulating large-scale atomic systems presents significant computational challenges.

Purpose of the Study:

• To present GAPD, a novel GPU-accelerated code for atom-based polychromatic diffraction simulations.
• To enable direct, kinematics-based simulations of X-ray/electron diffraction for large atomic systems.

Main Methods:

• Implementation of GPU parallel computation using real- and reciprocal-space decompositions.
• Development of a code capable of handling mono-/polychromatic beams and arbitrary detector geometries.
• Simulation of ultralarge systems (up to ~5 billion atoms).

Main Results:

• Successful direct simulations of reciprocal lattice nodes for ultralarge systems.
• Generation of diffraction patterns for single-crystal and polycrystalline configurations using various X-ray beam types.
• Validation and benchmarking of the GAPD code's performance and accuracy.

Conclusions:

• GAPD provides an efficient and scalable solution for simulating diffraction from large atomic systems.
• The code facilitates detailed analysis of diffraction patterns, aiding in materials characterization.
• GAPD supports diverse experimental configurations, including synchrotron undulator sources.