arXiv subjects

Jianmin Zhang

Publications and source records attributed to Jianmin Zhang.

12 recordsLinked to original sources

arXiv2026

NeuroPDE+: A Scalable Neuromorphic PDE Accelerator Based on Spintronic and Ferroelectric Devices

The pursuit of high-performance PDE solvers rests on three fundamental challenges: (i) the curse of dimensionality in kinetic and financial equations, (ii) the poor extrapolation of purely data-driven surrogates, and (iii) the widening gap between algorithm design and hardware specialization. To overcome these challenges, we present NeuroPDE+, a scalable neuromorphic PDE solver design based on spintronic and ferroelectric devices for accelerating PDE solutions. NeuroPDE+ consists of two dedicated units: a diffusion tracking unit (DTU), which emulates random walks on Markov chains through activations between hardware neurons, and a scattering tracking unit (STU), which samples non-local jumps via a multi-level probability tree. System-level simulations suggest that NeuroPDE+ achieves a squared error below 1e-2 in steady-state heat equation and particle transport problems. Simulation results further indicate that the DTU achieves up to a 315x performance gain over previous neuromorphic processors, and that the STU achieves a 1000x speedup compared to a general-purpose CPU. Co-designing algorithm and hardware with intrinsic stochasticity and non-volatile in-memory computing, NeuroPDE+ preliminarily explores a new paradigm for efficient and scalable neuromorphic PDE solvers. This approach could pave the way for probabilistic computing architectures in large-scale scientific simulations.

cs.AR↗

arXiv2026

C2P-Cache: Scalable GPU L1 Cache Sharing via Concurrent Candidate Pruning

Modern GPUs rely on private per-SM L1 caches and a shared L2 cache, but this organization obscures cross-SM reuse: an L1 miss is typically forwarded to L2 even when the requested line already resides in a peer L1 cache, leading to redundant L2 access. Prior GPU L1-sharing designs attempt to recover such reuse through exact or broad remote-hit searches, which become increasingly difficult to scale and can interfere with the critical L1 miss path under high concurrency. %miss handling as more caches participate and more misses arrive concurrently. We observe that eliminating redundant L2 accesses does not require exact, chip-wide knowledge of private L1 contents. Instead, it requires only sufficient visibility to sharply narrow down a small set of candidate caches, leaving exact confirmation to a much smaller number of L1s. Based on this insight, we propose C2P-Cache, a scalable GPU L1-sharing mechanism that transforms remote-hit discovery from a chip-wide exact search problem into a lightweight filtering-and-confirmation process. C2P-Cache maintains compact Bloom-filter-based snapshots of private L1 tags, performs parallel chip-wide candidate filtering, and selectively probes only a small number of likely peer caches. To sustain high concurrency, C2P-Cache organizes filtering as bit-sliced matching over a banked and replicated snapshot matrix, enabling efficient, parallel processing of many concurrent misses without interfering with normal L1 accesses. Across a wide range of GPU workloads, C2P-Cache improves instructions per cycle (IPC) by up to 49.7\% and by 23.5\% on average for applications with high remote-L1 reuse and strong sensitivity to L2 latency, demonstrating that lightweight, scalable filtering can effectively unlock cross-SM reuse with modest overhead.

cs.AR↗

arXiv2025

Spin-NeuroMem: A Low-Power Neuromorphic Associative Memory Design Based on Spintronic Devices

Biologically-inspired computing models have made significant progress in recent years, but the conventional von Neumann architecture is inefficient for the large-scale matrix operations and massive parallelism required by these models. This paper presents Spin-NeuroMem, a low-power circuit design of Hopfield network for the function of associative memory. Spin-NeuroMem is equipped with energy-efficient spintronic synapses which utilize magnetic tunnel junctions (MTJs) to store weight matrices of multiple associative memories. The proposed synapse design achieves as low as 17.4% power consumption compared to the state-of-the-art synapse designs. Spin-NeuroMem also encompasses a novel voltage converter with a 53.3% reduction in transistor usage for effective Hopfield network computation. In addition, we propose an associative memory simulator for the first time, which achieves a 5Mx speedup with a comparable associative memory effect. By harnessing the potential of spintronic devices, this work paves the way for the development of energy-efficient and scalable neuromorphic computing systems.

cs.AR↗

arXiv2025

NeuroPDE: A Neuromorphic PDE Solver Based on Spintronic and Ferroelectric Devices

In recent years, new methods for solving partial differential equations (PDEs) such as Monte Carlo random walk methods have gained considerable attention. However, due to the lack of hardware-intrinsic randomness in the conventional von Neumann architecture, the performance of PDE solvers is limited. In this paper, we introduce NeuroPDE, a hardware design for neuromorphic PDE solvers that utilizes emerging spintronic and ferroelectric devices. NeuroPDE incorporates spin neurons that are capable of probabilistic transmission to emulate random walks, along with ferroelectric synapses that store continuous weights non-volatilely. The proposed NeuroPDE achieves a variance of less than 1e-2 compared to analytical solutions when solving diffusion equations, demonstrating a performance advantage of 3.48x to 315x speedup in execution time and an energy consumption advantage of 2.7x to 29.8x over advanced CMOS-based neuromorphic chips. By leveraging the inherent physical stochasticity of emerging devices, this study paves the way for future probabilistic neuromorphic computing systems.

cs.AR↗

arXiv2024

RHS-TRNG: A Resilient High-Speed True Random Number Generator Based on STT-MTJ Device

High-quality random numbers are very critical to many fields such as cryptography, finance, and scientific simulation, which calls for the design of reliable true random number generators (TRNGs). Limited by entropy source, throughput, reliability, and system integration, existing TRNG designs are difficult to be deployed in real computing systems to greatly accelerate target applications. This study proposes a TRNG circuit named RHS-TRNG based on spin-transfer torque magnetic tunnel junction (STT-MTJ). RHS-TRNG generates resilient and high-speed random bit sequences exploiting the stochastic switching characteristics of STT-MTJ. By circuit/system co-design, we integrate RHS-TRNG into a RISC-V processor as an acceleration component, which is driven by customized random number generation instructions. Our experimental results show that a single cell of RHS-TRNG has a random bit generation speed of up to 303 Mb/s, which is the highest among existing MTJ-based TRNGs. Higher throughput can be achieved by exploiting cell-level parallelism. RHS-TRNG also shows strong resilience against PVT variations thanks to our designs using bidirectional switching currents and dual generator units. In addition, our system evaluation results using gem5 simulator suggest that the system equipped with RHS-TRNG can achieve 3.4-12x higher performance in speeding up option pricing programs than software implementations of random number generation.

cs.AR↗

arXiv2023

Decoding Chinese phonemes from intracortical brain signals with hyperbolic-space neural representations

Speech brain-computer interfaces (BCIs), which translate brain signals into spoken words or sentences, have shown significant potential for high-performance BCI communication. Phonemes are the fundamental units of pronunciation in most languages. While existing speech BCIs have largely focused on English, where words contain diverse compositions of phonemes, Chinese Mandarin is a monosyllabic language, with words typically consisting of a consonant and a vowel. This feature makes it feasible to develop high-performance Mandarin speech BCIs by decoding phonemes directly from neural signals. This study aimed to decode spoken Mandarin phonemes using intracortical neural signals. We observed that phonemes with similar pronunciations were often represented by inseparable neural patterns, leading to confusion in phoneme decoding. This finding suggests that the neural representation of spoken phonemes has a hierarchical structure. To account for this, we proposed learning the neural representation of phoneme pronunciation in a hyperbolic space, where the hierarchical structure could be more naturally optimized. Experiments with intracortical neural signals from a Chinese participant showed that the proposed model learned discriminative and interpretable hierarchical phoneme representations from neural signals, significantly improving Chinese phoneme decoding performance and achieving state-of-the-art. The findings demonstrate the feasibility of constructing high-performance Chinese speech BCIs based on phoneme decoding.

cs.HC↗

arXiv2022

Dynamic Ensemble Bayesian Filter for Robust Control of a Human Brain-machine Interface

Objective: Brain-machine interfaces (BMIs) aim to provide direct brain control of devices such as prostheses and computer cursors, which have demonstrated great potential for mobility restoration. One major limitation of current BMIs lies in the unstable performance in online control due to the variability of neural signals, which seriously hinders the clinical availability of BMIs. Method: To deal with the neural variability in online BMI control, we propose a dynamic ensemble Bayesian filter (DyEnsemble). DyEnsemble extends Bayesian filters with a dynamic measurement model, which adjusts its parameters in time adaptively with neural changes. This is achieved by learning a pool of candidate functions and dynamically weighting and assembling them according to neural signals. In this way, DyEnsemble copes with variability in signals and improves the robustness of online control. Results: Online BMI experiments with a human participant demonstrate that, compared with the velocity Kalman filter, DyEnsemble significantly improves the control accuracy (increases the success rate by 13.9% and reduces the reach time by 13.5% in the random target pursuit task) and robustness (performs more stably over different experiment days). Conclusion: Our results demonstrate the superiority of DyEnsemble in online BMI control. Significance: DyEnsemble frames a novel and flexible framework for robust neural decoding, which is beneficial to different neural decoding applications.

cs.LG↗

arXiv2020

Laser-induced damage thresholds of ultrathin targets and their constrain on laser contrast in laser-driven ion acceleration experiments

Single-shot laser-induced damage threshold (LIDT) measurements of multi-type free-standing ultrathin foils were performed in vacuum environment for 800 nm laser pulses with durations τ ranging from 50 fs to 200 ps. Results show that the laser damage threshold fluences (DTFs) of the ultrathin foils are significantly lower than those of corresponding bulk materials. Wide band gap dielectric targets such as SiN and formvar have larger DTFs than those of semiconductive and conductive targets by 1-3 orders of magnitude depending on the pulse duration. The damage mechanisms for different types of targets are studied. Based on the measurement, the constrain of the LIDTs on the laser contrast is discussed.

physics.plasm-ph↗

arXiv2020

Energy Efficiency Optimization for NOMA UAV Network with Imperfect CSI

Unmanned aerial vehicles (UAVs) are developing rapidly owing to flexible deployment and access services as air base stations. However, the channel errors of low-altitude communication links formed by mobile deployment of UAVs cannot be ignored. And the energy efficiency of the UAVs communication with imperfect channel state information (CSI) hasnt been well studied yet. Therefore, we focus on system performance optimization in non-orthogonal multiple access (NOMA) UAV network considering imperfect CSI between the UAV and users. A suboptimal resource allocation scheme including user scheduling and power allocation is designed for maximizing energy efficiency. Because of the nonconvexity of optimization function with an probability constraint for imperfect CSI, the original problem is converted into a non-probability problem and then decoupled into two convex subproblems. First, a user scheduling method is applied in the two-side matching of users and subchannels by the difference of convex programming. Then based on user scheduling, the energy efficiency in UAV cells is optimized through a suboptimal power allocation algorithm by successive convex approximation method. The simulation results prove that the proposed algorithm is effective compared with existing resource allocation schemes.

cs.IT↗

arXiv2018

On the relationship between the diffuse reflection and bounce-back boundary condition in the continuum limit

In this work, we show that the widely used bounce-back boundary condition is an incomplete form of the diffuse reflection boundary condition at the continuum limit for lattice Boltzmann simulations. By utilizing this fact, we can force the diffuse reflection scheme to work at its continuum limit so that the no-slip boundary condition can be implemented without any non-physical slip velocity error being induced by the standard bounce-back scheme. The revised boundary formulation is validated numerically by solving a pressure-driven channel flow, a lid-driven cavity flow and channel flow around a square cylinder.

physics.flu-dyn↗

arXiv2018

Discrete Boltzmann model of shallow water equations with polynomial equilibria

A type of discrete Boltzmann model for simulating shallow water flows is derived by using the Hermite expansion approach. Through analytical analysis, we study the impact of truncating distribution function and discretizing particle velocity space. It is found that the convergence behavior of expansion is nontrivial while the conservation laws are naturally satisfied. Moreover, the balance of source terms and flux terms for steady solutions is not sacrificed. Further numerical validations show that the capability of simulating supercritical flows is enhanced by employing higher order expansion and quadrature.

physics.flu-dyn↗

arXiv2018

Towards a Neural Network Approach to Abstractive Multi-Document Summarization

Till now, neural abstractive summarization methods have achieved great success for single document summarization (SDS). However, due to the lack of large scale multi-document summaries, such methods can be hardly applied to multi-document summarization (MDS). In this paper, we investigate neural abstractive methods for MDS by adapting a state-of-the-art neural abstractive summarization model for SDS. We propose an approach to extend the neural abstractive model trained on large scale SDS data to the MDS task. Our approach only makes use of a small number of multi-document summaries for fine tuning. Experimental results on two benchmark DUC datasets demonstrate that our approach can outperform a variety of baseline neural models.

cs.CL↗