Grain - A Java Analysis Framework for Total Data Readout

Grain is a data analysis framework developed to be used with the novel Total Data Readout data acquisition system. In Total Data Readout all the electronics channels are read out asynchronously in singles mode and each data item is timestamped. Event building and analysis has to be done entirely in the software post-processing the data stream. A flexible and efficient event parser and the accompanying software framework have been written entirely in Java. The design and implementation of the software are discussed along with experiences gained in running real-life experiments.

💡 Research Summary

The paper presents Grain, a Java‑based analysis framework designed to handle the data streams produced by the Total Data Readout (TDR) acquisition system. Unlike conventional trigger‑based DAQ, TDR reads every detector channel asynchronously, timestamps each datum with 10 ns precision, and delivers a single time‑ordered stream that contains no built‑in event structure. Consequently, event building, veto handling, and pile‑up rejection must be performed entirely in software. Grain addresses these challenges through a modular, multithreaded architecture.

The processing chain begins with stream filtering, where software identifies vetoed and pile‑up signals. Bit‑pattern registers and ADC channels are cross‑checked to either discard or flag unwanted data. Two event parsers are provided: a decay/RDT parser that uses any hit in a predefined implantation detector as a trigger, and a stand‑alone parser that triggers on a multiplicity condition across a set of detectors. Both parsers define a time slice around the trigger using three parameters – trigger channel, offset, and width – and extract all data within that window from a ring‑buffer. The extracted slice is then routed into sub‑detector objects, forming a hierarchical event object that can be further correlated.

Grain’s framework exploits modern multi‑core CPUs. The graphical user interface (GUI) runs on a master thread, while separate worker threads perform sorting, parsing, and histogramming, communicating via FIFO buffers. Dynamic class loading enables users to write custom sorters in Java, providing great flexibility. Histograms and n‑tuples are managed through the JAIDA library, adhering to the AIDA standard; a custom multidimensional binner was added to support nuclear‑physics‑specific data structures. Output files are stored in AIDA‑compatible XML, allowing interoperability with ROOT, Radware, and other analysis tools.

Performance tests on a dual‑core AMD system showed that Grain can sustain data rates of up to 3.2 MB s⁻¹ (≈400 kHz) from the TDR system. When raw histogramming is disabled, throughput reaches 44 MB s⁻¹ for simple γ‑γ experiments and 28 MB s⁻¹ for triple‑coincidence measurements. Profiling revealed that the event parser consumes about 65 % of total CPU time, with the ring‑buffer search being the primary bottleneck. Nevertheless, multithreading mitigates this limitation by keeping other tasks on separate cores, preserving overall throughput.

Since 2002, Grain has been employed in over fifty experiments, ranging from heavy‑element γ‑spectroscopy to recoil‑decay tagging (RDT) studies in the A≈100 region and the development of β‑tagging techniques. The choice of Java provided platform independence, a rich set of GUI components, and robust networking capabilities, which together facilitated easy installation on Windows, Linux, and macOS platforms. The authors conclude that Grain offers a flexible, efficient, and user‑friendly solution for trigger‑less data acquisition, demonstrating that modern Java runtimes can meet the performance demands of high‑rate nuclear physics experiments.