High-Performance Multi-Mode Ptychography Reconstruction on Distributed GPUs

TL;DR

This paper presents a GPU-accelerated multi-mode ptychography reconstruction algorithm that significantly reduces processing time from hours to about a minute, enabling real-time imaging applications.

Contribution

The authors develop a distributed GPU implementation of the multi-mode difference map algorithm for ptychography, achieving high speedup and near-constant weak scaling.

Findings

Speedup of 10 to 1000 times depending on data size.

Reconstruction time reduced from hours to about 1 minute.

Supports real-time data processing and visualization.

Abstract

Ptychography is an emerging imaging technique that is able to provide wavelength-limited spatial resolution from specimen with extended lateral dimensions. As a scanning microscopy method, a typical two-dimensional image requires a number of data frames. As a diffraction-based imaging technique, the real-space image has to be recovered through iterative reconstruction algorithms. Due to these two inherent aspects, a ptychographic reconstruction is generally a computation-intensive and time-consuming process, which limits the throughput of this method. We report an accelerated version of the multi-mode difference map algorithm for ptychography reconstruction using multiple distributed GPUs. This approach leverages available scientific computing packages in Python, including mpi4py and PyCUDA, with the core computation functions implemented in CUDA C. We find that interestingly even with MPI collective communications, the weak scaling in the number of GPU nodes can still remain nearly constant. Most importantly, for realistic diffraction measurements, we observe a speedup ranging from a factor of $10$ to $10^3$ depending on the data size, which reduces the reconstruction time remarkably from hours to typically about 1 minute and is thus critical for real-time data processing and visualization.