arXiv subjects

Wei-Chen Chen

Publications and source records attributed to Wei-Chen Chen.

5 recordsLinked to original sources

arXiv2026

Cross-Domain Acceleration of Open Modification Search: From Commodity Platforms to Emerging Memory and Storage Devices

Open modification search (OMS) in mass spectrometry (MS) is a data-intensive workload whose performance is dominantly limited by reference data movement rather than computation. Prior OMS accelerators have largely been evaluated in isolation, making it difficult to understand system-level trade-offs across platforms. This paper presents the first workload-driven, cross-platform survey of accelerators for MS search by studying not only commodity platforms, but also emerging memory- and storage-centric architectures, including GPUs, near-storage FPGAs, DRAM near-memory processing, ReRAM/PCM in-memory processing, and 3D NAND/FeNAND in-storage processing, under consistent algorithmic and accuracy assumptions. Leveraging a binary hyperdimensional computing (HDC)-based OMS formulation that reduces similarity evaluation to lightweight bitwise primitives and tolerates device-level non-idealities, we enable a robust execution on memory-centric architectures despite device-level non-idealities and limited computing capability. Overall, this study identifies memory- and storage-centric architectures as a key architectural breakthrough for large-scale, high-speed search acceleration, delivering up to >100x speedup and >40,000x improvement in energy efficiency.

cs.AR↗

arXiv2025

Low-power Spike-based Wearable Analytics on RRAM Crossbars

This work introduces a spike-based wearable analytics system utilizing Spiking Neural Networks (SNNs) deployed on an In-memory Computing engine based on RRAM crossbars, which are known for their compactness and energy-efficiency. Given the hardware constraints and noise characteristics of the underlying RRAM crossbars, we propose online adaptation of pre-trained SNNs in real-time using Direct Feedback Alignment (DFA) against traditional backpropagation (BP). Direct Feedback Alignment (DFA) learning, that allows layer-parallel gradient computations, acts as a fast, energy & area-efficient method for online adaptation of SNNs on RRAM crossbars, unleashing better algorithmic performance against those adapted using BP. Through extensive simulations using our in-house hardware evaluation engine called DFA_Sim, we find that DFA achieves upto 64.1% lower energy consumption, 10.1% lower area overhead, and a 2.1x reduction in latency compared to BP, while delivering upto 7.55% higher inference accuracy on human activity recognition (HAR) tasks.

cs.ET↗

arXiv2024

Efficient Open Modification Spectral Library Searching in High-Dimensional Space with Multi-Level-Cell Memory

Open Modification Search (OMS) is a promising algorithm for mass spectrometry analysis that enables the discovery of modified peptides. However, OMS encounters challenges as it exponentially extends the search scope. Existing OMS accelerators either have limited parallelism or struggle to scale effectively with growing data volumes. In this work, we introduce an OMS accelerator utilizing multi-level-cell (MLC) RRAM memory to enhance storage capacity by 3x. Through in-memory computing, we achieve up to 77x faster data processing with two to three orders of magnitude better energy efficiency. Testing was done on a fabricated MLC RRAM chip. We leverage hyperdimensional computing to tolerate up to 10% memory errors while delivering massive parallelism in hardware.

cs.AR↗

arXiv2023

A practical model-based segmentation approach for improved activation detection in single-subject functional Magnetic Resonance Imaging studies

Functional Magnetic Resonance Imaging (fMRI) maps cerebral activation in response to stimuli but this activation is often difficult to detect, especially in low-signal contexts and single-subject studies. Accurate activation detection can be guided by the fact that very few voxels are, in reality, truly activated and that these voxels are spatially localized, but it is challenging to incorporate both these facts. We address these twin challenges to single-subject and low-signal fMRI by developing a computationally feasible and methodologically sound model-based approach, implemented in the R package MixfMRI, that bounds the a priori expected proportion of activated voxels while also incorporating spatial context. An added benefit of our methodology is the ability to distinguish voxels and regions having different intensities of activation. Our suggested approach is evaluated in realistic two- and three-dimensional simulation experiments as well as on multiple real-world datasets. Finally, the value of our suggested approach in low-signal and single-subject fMRI studies is illustrated on a sports imagination experiment that is often used to detect awareness and improve treatment in patients in persistent vegetative state (PVS). Our ability to reliably distinguish activation in this experiment potentially opens the door to the adoption of fMRI as a clinical tool for the improved treatment and therapy of PVS survivors and other patients.

stat.AP↗

arXiv2017

Parallel Statistical Computing with R: An Illustration on Two Architectures

To harness the full benefit of new computing platforms, it is necessary to develop software with parallel computing capabilities. This is no less true for statisticians than for astrophysicists. The R programming language, which is perhaps the most popular software environment for statisticians today, has many packages available for parallel computing. Their diversity in approach can be difficult to navigate. Some have attempted to alleviate this problem by designing common interfaces. However, these approaches offer limited flexibility to the user; additionally, they often serve as poor abstractions to the reality of modern hardware, leading to poor performance. We give a short introduction to two basic parallel computing approaches that closely align with hardware reality, allow the user to understand its performance, and provide sufficient capability to fully utilize multicore and multinode environments. We illustrate both approaches by working through a simple example fitting a random forest model. Beginning with a serial algorithm, we derive two parallel versions. Our objective is to illustrate the use of multiple cores on a single processor and the use of multiple processors in a cluster computer. We discuss the differences between the two versions and how the underlying hardware is used in each case.

stat.CO↗