Results 91 to 100 of about 32,736 (268)

Physics‐Grounded Probabilistic Bits for Hardware‐Efficient Intelligent Inference and Optimization

Advanced Intelligent Systems, EarlyView.
Si–SiNx interface traps are harnessed as a complementary metal–oxide–semiconductor‐compatible source of controllable randomness for probabilistic bits. Pulse‐width‐programmed stochastic capture converts nanoscale defect dynamics into Boltzmann‐consistent binary outputs, while a physics‐based Simulation Program with Integrated Circuit Emphasis model ...
Dokyoung Lee, Jun‐Young Park, Joon‐Kyu Han, Sungho Kim   +3 more
wiley   +1 more source

The collective computing model

Journal of Computer Science and Technology, 2000
The parallel computing model used in this paper, the Collective Computing Model (CCM), is a variant of the well-known Bulk Synchronous Parallel (BSP) model. The synchronicity imposed by the BSP model restricts the set of available algorithms and prevents
Jesús Alberto Gonzalez, Coromoto León, María Fabiana Piccoli, Alicia Marcela Printista, José Luis Roda García, Casiano Rodríguez, Francisco de Sande   +6 more
doaj  

Energy‐Aware Perturbation Optimization for Memristor‐Array Convolutional Neural Networks

Advanced Intelligent Systems, EarlyView.
Memristor‐array inference becomes more energy efficient when layer inputs are reshaped before computation. Sinusoidal perturbation encoding with dual‐threshold screening reduces active voltage pulses and contracts ADC input‐current ranges, jointly lowering crossbar and peripheral energy while preserving accuracy across hardware MNIST validation, deep ...
Ao Xu, Lun Lu, Youpeng Wu, Yao Guo, Yi Li, Yinan Wang, Qingjiang Li   +6 more
wiley   +1 more source

Single‐Step Analog In‐Memory Matrix Computation in Three‐Dimensional Circuits

Advanced Intelligent Systems, EarlyView.
A wooden model illustrates a 3D memristive array designed for single‐step kernel convolutions. It shows how two‐dimensional input data are mapped into the 3D array and convolved in parallel with stored kernels. Color‐coded connection points represent programmed memory‐element values, with each color corresponding to a specific kernel weight.
Alireza Jaberi Rad, Koushik Kumar, Fatemeh Kiani, Yi Huang, Ali Abdel‐Maksoud, Zheshun Xiong, Yunzhi Ling, Wuyu Zhao, Qiangfei Xia   +8 more
wiley   +1 more source

A Scalable and Resource‐Efficient Pipelined p‐Computer for Probabilistic Ising Machines

Advanced Intelligent Systems, EarlyView.
(a) Block diagram of the portfolio optimization problem: given M assets, the goal is to determine the optimal weights w that maximize the expected return (based on the mean historical assets return u), while minimizing the risk, quantified by the assets covariance matrix S.
Deborah Volpe, Eleonora Raimondo, Andrea Grimaldi, Pedram Khalili Amiri, Stefano Chiappini, Anna Giordano, Mario Carpentieri, Hyunsoo Yang, Massimo Chiappini, Giovanni Finocchio   +9 more
wiley   +1 more source

Sociodemographic Factors Associated with Hours Worked by Primary Carers in Australia

Australian Journal of Social Issues, EarlyView.
ABSTRACT Primary caregivers constitute a major unpaid workforce in Australia. The aim of this study was to determine the sociodemographic factors that are associated with carer workloads. Multinomial logistic regression modelling was applied to the nation‐wide Australian Government survey.
Andrew J. Hamilton
wiley   +1 more source

Parallelizing Parallel Programs [PDF]

Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques, 2020
Roberto Castañeda Lozano, Murray Cole, Björn Franke   +2 more
openaire   +2 more sources

The Circular Economy—A New Resilience Paradigm?

Business Strategy and the Environment, EarlyView.
ABSTRACT The circular economy (CE) has traditionally been positioned as a central instrument of the sustainability paradigm. Yet escalating geopolitical volatility and supply chain fragility have elevated resilience to a parallel strategic imperative, and the conceptual integration of CE and resilience remains underexplored.
Julian Kirchherr, Felix Carl Schultz
wiley   +1 more source

Model Parallelism With Subnetwork Data Parallelism

CoRR
Pre-training large neural networks at scale imposes heavy memory demands on accelerators and often requires costly communication. We introduce Subnetwork Data Parallelism (SDP), a distributed training framework that partitions a model into structured subnetworks trained across workers without exchanging activations.
Singh, Vaibhav, Khalid, Zafir, Cagnasso, Pietro, Oyallon, Edouard, Belilovsky, Eugene   +4 more
openaire   +3 more sources

Comparison Between the Acurate Neo and the Acurate Neo 2 Transcatheter Heart Valve System: Impact of Midframe Under‐Expansion

Catheterization and Cardiovascular Interventions, EarlyView.
ABSTRACT Background Transcatheter aortic valve implantation (TAVI) with the Acurate Neo has been linked to a significant incidence of paravalvular leakage (PVL), addressed by the Acurate Neo 2 incorporating an active sealing skirt. Recent findings suggest that midframe under‐expansion is associated with worse outcomes.
Lisa Crusius, Thaddäus Ast, Stephan Haussig, Jean‐Honoré Steul, Felix J. Woitek, Peter Sick, Tomas Madej, Utz Kappert, Axel Linke, Norman Mangner   +9 more
wiley   +1 more source