arXiv subjects

M. Cherukara

Publications and source records attributed to M. Cherukara.

3 recordsLinked to original sources

arXiv2026

On-chip Bragg peak extraction from MHz frame rate X-ray detectors using a cellular automaton architecture

High-frame rate pixel detectors can produce data volumes that exceed available off-chip bandwidth, yet in many applications only a sparse subset of each frame carries relevant information. Spatially localized events, including diffraction peaks in crystallography, particle hits in tracking detectors, fluorescence spots in biological imaging, and other applications, all require that clusters of above-threshold pixels be identified and extracted from an otherwise featureless background. Conventionally this is performed in software algorithms such as connected-component labeling on full frames, but at MHz frame rates the resulting throughput becomes prohibitive. We present a lightweight hardware architecture that performs peak localization and patch extraction directly in the sensor silicon, transmitting only small pixel patches rather than complete frames. FPGA-based testing on an AMD Alveo V80 validated the synthesizability and timing closure of the peak-finding module in real hardware. The design replaces global connected-component labeling with a cellular automaton that uses purely local, fixed-iteration neighborhood operations, eliminating the label storage and equivalence-resolution logic that make conventional approaches impractical on-chip. We validate the architecture on X-ray Bragg peak detection for far-field high-energy diffraction microscopy and show that every peak found by a software reference is recovered, with equivalent downstream reconstructions. The architecture sustains several-hundred-kHz frame rates in 130nm CMOS and exceeds 1MHz in 28nm.

physics.ins-det↗

arXiv2023

Spontaneous supercrystal formation during a strain-engineered metal-insulator transition

Mott metal-insulator transitions possess electronic, magnetic, and structural degrees of freedom promising next generation energy-efficient electronics. We report a previously unknown, hierarchically ordered state during a Mott transition and demonstrate correlated switching of functional electronic properties. We elucidate in-situ formation of an intrinsic supercrystal in a Ca2RuO4 thin film. Machine learning-assisted X-ray nanodiffraction together with electron microscopy reveal multi-scale periodic domain formation at and below the film transition temperature (TFilm ~ 200-250 K) and a separate anisotropic spatial structure at and above TFilm. Local resistivity measurements imply an intrinsic coupling of the supercrystal orientation to the material's anisotropic conductivity. Our findings add an additional degree of complexity to the physical understanding of Mott transitions, opening opportunities for designing materials with tunable electronic properties.

cond-mat.mtrl-sci↗

arXiv2023

Neural Network Methods for Radiation Detectors and Imaging

Recent advances in image data processing through machine learning and especially deep neural networks (DNNs) allow for new optimization and performance-enhancement schemes for radiation detectors and imaging hardware through data-endowed artificial intelligence. We give an overview of data generation at photon sources, deep learning-based methods for image processing tasks, and hardware solutions for deep learning acceleration. Most existing deep learning approaches are trained offline, typically using large amounts of computational resources. However, once trained, DNNs can achieve fast inference speeds and can be deployed to edge devices. A new trend is edge computing with less energy consumption (hundreds of watts or less) and real-time analysis potential. While popularly used for edge computing, electronic-based hardware accelerators ranging from general purpose processors such as central processing units (CPUs) to application-specific integrated circuits (ASICs) are constantly reaching performance limits in latency, energy consumption, and other physical constraints. These limits give rise to next-generation analog neuromorhpic hardware platforms, such as optical neural networks (ONNs), for high parallel, low latency, and low energy computing to boost deep learning acceleration.

physics.ins-det↗