Latest Articles for Computer Architecture

A new approach improves deep learning simulation in hardware design.

Introducing FLAASH, a system designed to speed up tensor contractions in machine learning.

Research highlights fast switching and low energy use of VO memristors.

A new method detects data races with improved accuracy and low overhead.

This study examines methods for efficient GPU resource management under power constraints.

A method to automate and optimize instruction selection for various computer architectures.

A new B+-tree structure enhances data management in disaggregated memory systems.

Efficient execution of transformer models on an open-source RISC-V platform.

New methods improve efficiency in digital circuit design using ACD.

A look at kernel safety methods to combat advanced attacks like Spectre.

FRED enhances communication for faster and more efficient deep learning training.

New Ising machine model improves optimization problem solving and computational speed.

A new method improves verification of processor security against speculative execution vulnerabilities.

This article examines RVV's impact on ANN algorithm performance for faster data processing.

Triangular Input Movement enhances efficiency in convolutional neural networks.

A new toolchain improves memory safety for WASM applications using modern hardware.

Skeia tackles shadow branches to enhance instruction fetching efficiency.

New techniques reduce energy use in deep learning while maintaining accuracy.

This paper addresses control-flow leakage threats and proposes a new hardware-software approach.

Exploring the shift toward GPU-centered methods for better computing performance.

A new method predicts dataflow strategies for better GNN processing.

Explore the rising popularity of ARM chips and their impact on technology.

Discover Massimult, a fresh architecture for faster and efficient computing.

MQMS transforms GPU-SSD systems for faster data processing.

Plaid redefines computing efficiency by aligning resources for peak performance with minimal energy use.

SparseMap streamlines data management for efficient neural network processing.