FPGA-Based Real-Time Image Manipulation and Advanced Data Acquisition for 2-D-XRAY Detectors

Abstract

Scientific experiments rely on some type of measurements that provides the required data to extract aimed information or conclusions. Data production and analysis are therefore essential components at the heart of any scientific experimental application.Traditionally, efforts on detector development for photon sources have focused on the properties and performance of the detection front-ends. In many cases, the data acquisition chain as well as data processing, are treated as a complementary component of the detector system and added at a late stage of the project. In most of the cases, data processing tasks are entrusted to CPUs; achieving thus the minimum bandwidth requirements and kept hardware relatively simple in term of functionalities. This also minimizes design effort, complexity and implementation cost.This approach is changing in the last years as it does not fit new high-performance detectors; FPGA and GPUs are now used to perform complex image manipulation tasks such as image reconstruction, image rotation, accumulation, filtering, data analysis and many others. This frees up CPUs for simpler tasks.The objective of this paper is to present both the implementation of real time FPGA-based image manipulation techniques, as well as, the performance of the ESRF data acquisition platform called RASHPA, into the back-end board of the SMARTPIX photon-counting detector developed at the ESRF.

Evaluation of HPC Acceleration and Interconnect Technologies for High-Throughput Data Acquisition

“…The authors of [6] address the problem of acquiring data from 2D X-ray detectors (either images, regions of interest, metadata, or events) and delivering them to multiple backend computers by developing an RDMA over Ethernet protocol and designing a 100 G MAC component for Xilinx UltraScale+ FPGAs and then comparing the dedicated link with a standard RDMA over Converged Ethernet v2 (RoCEv2) protocol using commercial Mellanox adapters. Based on this, the work in [7] implements a real-time FPGA-based image manipulation system used in a backend board of a photon-counting detector as part of the RASHPA data acquisition platform at ESRF. The contribution in [8] addresses macromolecular crystallography, the dominant method for high-resolution structure determination of biomolecules, and specifically, it uses raw Ethernet (or zero-copy) transfers for low-latency data acquisition from gain adaptive detectors, which can be accomplished without the CPU in the receiving machines being involved in the transfers themselves.…”

“…At the same time, these applications pose unprecedented processing requirements for the vast amount of data being generated and transferred, increasingly leading to the adoption of accelerator-based platforms, e.g., using graphics processing units (GPUs) or field-programmable gate arrays (FPGAs) in order to meet stringent energy-efficiency requirements through some form of customized computing. Consequently, many very recent works have explored advanced solutions for improving the end-to-end data communication performance targeting accelerator-based platforms, particularly those based on customized FPGA accelerators, in the context of scientific data acquisition and scientific computing [3,[5][6][7][8][9].…”