Articles tagged with Computer Architecture
Researchers have developed a novel optical neural network architecture that achieves nonlinear optical computation by precisely controlling ultrashort pulse propagation in multimode fibers. This approach streamlines the need for energy-intensive digital processes, achieving comparable accuracy with significantly reduced parameters.
Incheon National University researchers developed a web 3.0 streaming architecture that reduces delay, improves user experience, and ensures transparency and fairness for real-time services. The proposed system uses Inter-Planetary file system (IPFS) to enable blockchain-based peer-to-peer data storage and caching.
Researchers extend spatially incoherent diffractive networks to perform complex-valued linear transformations with negligible error, opening up new applications in fields like autonomous vehicles. This breakthrough enables the encryption and decryption of complex-valued images using spatially incoherent diffractive networks.
Researchers will incorporate advanced semiconductor technologies and AI into a millimeter-wave radio system to increase bandwidth while reducing energy consumption. The project aims to save tens to hundreds of terawatt-hours of energy per year, contributing to climate change mitigation.
A new study using twisted magnets as computational medium has made brain-inspired computing more adaptable, reducing energy use and potential carbon emissions. The research found that by applying magnetic fields and changing temperature, physical properties of the materials can be adapted to suit different machine-learning tasks.
Researchers developed a novel photonic processor with adaptive neural connectivity, allowing for the creation of complex artificial neural networks. The system utilizes waveguide-coupled phase-change material to create almost 8,400 optical neurons that can adapt their connections through synaptic and structural plasticity.
Researchers develop integrated photonic-electronic hardware capable of processing three-dimensional (3D) data, doubling parallelism for AI tasks and significantly boosting energy efficiency. The new chip can process 100 electrocardiogram signals simultaneously with high accuracy, outperforming electronic processors.
The High Performance Data Facility Hub will provide researchers with unprecedented data management resources, accelerating scientific discovery through seamless access to large and complex datasets. The hub will be led by Jefferson Lab and partner with Lawrence Berkeley National Laboratory.
Researchers aim to develop programmable formal specification-based data stream processor and hardware monitor to enhance microchip security and prevent malfunctions. They will explore novel technologies for real-time monitoring of physical and biological systems, including signal patterns within computer chips.
Researchers at MIT have developed a novel superconducting qubit architecture that can perform operations between qubits with high accuracy, exceeding 99.9% for two-qubit gates and 99.99% for single-qubit gates. The new design utilizes fluxonium qubits, which have longer lifespans than traditional transmon qubits.
Rice University researchers have been awarded a 4-year, $1.2 million grant from the Department of Energy to evaluate different physical systems used to build quantum computers. The project aims to provide a framework for comparing the viability and computational potential of various approaches to building quantum computers.
Researchers at EPFL developed a novel system integrating 2D semiconductors and ferroelectric materials to create faster, more efficient electronics with brain-inspired operations. The technology enables significant energy reduction and advanced functionalities, including synaptic neuron function within the same device.
The WVU team evaluated Code Interpreter's features, finding it accessible to students but limited for scientists working with biological data. The plugin breaks down barriers for coding, but lacks internet access and parallel processing capabilities.
Researchers found a novel method to read data from CPUs by analyzing power consumption, dubbed Collide+Power. This attack consumes power and causes delays, allowing attackers to derive targeted data.
A breakthrough in photonic memory has been achieved, enabling fast volatile modulation and nonvolatile weight storage for rapid training of optical neural networks. The 5-bit photonic memory utilizes a low-loss PCM antimonite to achieve rapid response times and energy-efficient processing.
Lero is recruiting 16 top international post-doctoral researchers for a €2.9 million fellowship program focusing on privacy, trust, inclusion and fairness in software expertise. The program will provide discipline-specific skills training and enhance career development.
A new device from NIST scientists helps reduce noise in quantum computers by introducing a programmable toggle switch. This allows for more versatile quantum processors with clearer outputs and easier reprogramming, addressing long-standing challenges in quantum computing.
Ferroelectric materials like hafnia show promise for non-volatile random-access memory (RAM) due to their stability at high temperatures. Hafnia's unique properties, including the movement of oxygen vacancies, make it an attractive candidate for memristors that mimic brain-like computer architectures.
Optical memristors have the potential to transform high-bandwidth neuromorphic computing, machine learning hardware, and artificial intelligence. However, scalability is a significant challenge that needs to be addressed to unlock their full potential.
A new Danish research project, CP-SENS, aims to develop a digital twin platform for the mechanical and construction industries. The project will provide companies with access to intelligent IT systems tailored to their needs, enabling them to adopt digital twins without significant financial investment.
Researchers at North Carolina State University have developed a new methodology called Patch-to-Cluster attention (PaCa) that addresses the challenges of vision transformers. PaCa improves ViT's ability to identify, classify, and segment objects in images while reducing computational demands and enhancing model interpretability.
EPFL researchers have discovered a way to store and process data using magnetic waves, potentially solving the issue of energy-hungry computing technology. This approach enables non-volatile storage within the same system, reducing the need for separate processors and memory storage.
The TANGO project seeks to establish a symbiotic relationship between humans and machines, enabling effective and innovative AI systems that expand human reasoning and decision-making capabilities. Real-world use cases will be evaluated to assess the framework's potential impact on individuals and society.
The new architecture reduces physical qubits required for error correction to 10% of conventional architectures, enabling better performance than classical computers. This breakthrough accelerates progress toward practical quantum computing, with the aim of applying quantum computing applications to various societal issues.
Researchers developed memristors based on halogenated perovskite nanocrystals for more powerful and energy-efficient computing. Inspired by the human brain's synapses, these components combine data storage and processing, reducing energy consumption.
Researchers developed novel memristors with halide perovskite nanocrystals, enabling complex calculations similar to brain processes. The new memristors are faster, more energy-efficient, and easier to manufacture than predecessors.
Scientists from the University of Groningen develop complex oxide devices for energy-efficient computing, including magneto-electric spin-orbit and memristive devices. These materials have potential applications in novel computing architectures, such as random number generators.
Researchers have developed a new device that can effectively redistribute noise and reduce its impact on quantum measurements. By 'squeezing' the noise, they can make more accurate measurements, enabling faster and more precise quantum systems. The device has the potential to improve multi-qubit systems and metrological applications.
Researchers have developed a quantum computing architecture that enables directional photon emission, the first step toward extensible quantum interconnects. This breakthrough enables the creation of larger-scale devices by linking multiple processing modules along a common waveguide.
Researchers at Shinshu University demonstrate the transformation of isolated skyrmions into bimerons in a magnetic disk, showcasing a potential new operation for future computing architectures. The discovery opens up novel spintronic applications based on different topological spin textures.
Researchers at the University of Innsbruck have developed a new architecture for universal quantum computers using parity-based qubits. This design reduces the complexity of implementing complex algorithms while also offering hardware-efficient error correction.
Researchers at MIT have developed a new method that uses optics to accelerate machine-learning computations on low-power devices. By encoding model components onto light waves, data can be transmitted rapidly and computations performed quickly, leading to over a hundredfold improvement in energy efficiency.
Researchers propose a new solution called labeled von Neumann architecture (LvNA) to tackle challenges in cloud computing. LvNA incorporates label-powered control mechanisms to differentiate, isolate, and prioritize user-defined application requests, mitigating resource contentions and guaranteeing performance.
A University at Buffalo-led research team has been awarded a $5 million grant to develop digital tools that can help older adults recognize and protect themselves from online deceptions. The project, DART, aims to reduce online fraud among older adults, who lose billions of dollars each year due to scams.
The new computer chip uses a transistor-free design that eliminates data transfer time and minimizes energy consumption. It offers up to 100 times faster performance than conventional computing architectures, making it ideal for AI applications.
Researchers at Stanford University have created a new chip architecture called NeuRRAM that performs AI computing directly within memory, reducing energy consumption and increasing efficiency. The chip has been tested on various AI tasks and shown high accuracy rates.
The NeuRRAM chip demonstrates wide range of AI applications with equivalent accuracy while reducing energy consumption by up to 70% compared to traditional compute platforms. It also supports various neural network models and architectures, enabling diverse AI applications on edge devices.
Researchers at Columbia University have developed a new verification technology for the Arm Confidential Compute Architecture, demonstrating the first formal verification of a prototype. This breakthrough enables the creation of confidential computing architectures that can protect sensitive user data.
Researchers at Carnegie Mellon University are developing a new approach to harness the power of nanosatellites, collecting data insights while in orbit and reducing latency issues. This initiative will lay groundwork for innovative applications in fields like carbon mapping, traffic management, and precision agriculture.
Researchers designed a modular AI chip that can be easily upgraded by swapping out layers, reducing the need for new devices. The chip uses optical communication to transmit information between layers, enabling high versatility in edge computing applications.
Researchers at University of Virginia are developing long-lived sensors for the Internet of Things (IoT) to reduce maintenance costs and upgrade challenges. The goal is to create an ecosystem that enables sensors to adapt to changing environments, reducing electronic waste.
A team of researchers from Waseda University developed a novel solution to efficiently solve complex optimization problems using Ising machines. Their hybrid algorithm reduces residual energy and reaches more optimal results in shorter time, increasing the machine's applicability across industries and sustainability practices.
Researchers at North Carolina State University have developed FAXID, a hardware-based approach for detecting ransomware that is significantly faster than software-based methods. In proof-of-concept testing, FAXID demonstrated accuracy comparable to XGBoost but with speeds up to 65.8 times faster.
Researchers have developed a new encryption technique that leverages hardware and software to improve file system security for next-generation non-volatile memories. This approach allows for faster performance than existing software security technologies, making it suitable for large data centers and cloud systems.
Researchers at Washington State University have demonstrated a way to make memristors using honey, which can mimic the work of human synapses and process data in memory. The honey memristor chips could lead to the development of neuromorphic computing systems that function like the human brain.
Researchers developed a new reading method for SOT-RAMs that can nullify the readout disturbance, reducing it by at least 10 times. The method involves creating a bi-directional read path, cancelling out the disturbances produced by spin currents.
GIST researchers propose a new strategy for crime prevention using artificial intelligence, trained on a large-scale dataset of deviant incident reports and corresponding images. The model, called DevianceNet, can accurately classify and detect deviant places, making it a useful tool in urban safety development.
Researchers developed DAGguise, a scheme that shapes memory requests into a predefined pattern to prevent contention attacks and enable faster computation. The technique represents programs' memory access requests as a graph, where each request is stored in a node, and the edges are time dependencies between requests.
Scientists have developed a new approach to mimic synapse functions using magneto-ionics, which drastically decreases power consumption. The study reveals that thinner films of cobalt oxide exhibit faster magnetization generation and can emulate 'learning' and 'forgetting' functionalities.
A study by the University of Texas at Austin found that software development teams given greater autonomy are more productive and have higher customer satisfaction rates. The researchers tested 461 projects over 50 months and found a 39% increase in value added for autonomous teams compared to traditional teams.
A team of scientists from Gwangju Institute of Science and Technology developed a deep learning-based approach to predict SC2 battle outcomes by considering army composition and terrain type. The proposed model leveraged parameter sharing, enabling it to analyze complex factors accurately and make predictions.
A new AI algorithm, APOLLO, accurately predicts microprocessor power consumption by analyzing just 100 signals out of millions, offering potential to improve efficiency and develop new processors. The technique has been validated on high-performance microprocessors and could help designers inform future chip design.
Washington University researchers have designed a new processing-in-memory (PIM) circuit that can increase PIM computing's performance by orders of magnitude. The circuit uses resistive random-access memory PIM, allowing for analog computations and eliminating the need for digital conversions.
The University of Texas at Arlington has received a $1.5 million grant from the US Department of Education to support domestic doctoral students specializing in Internet-of-Things study and research. The project aims to increase diversity among computer science researchers, with a focus on women and minorities.
Fred Martin, a UMass Lowell computer scientist, is being recognized for developing an educational framework that introduces robotics and AI to K-12 students. His work has guided the development of primary and secondary AI education initiatives worldwide.
A new brain-inspired computing device harnesses randomness to improve AI and machine learning performance. The device, called hetero-memristor, emulates the stochastic properties of neurons, allowing for more efficient optimization and reduced power consumption.
Scientists at Aarhus University are working on a nano-sized brain-inspired computer that can harvest its own energy, making it the smallest and most efficient AI system yet. The project aims to reduce power consumption by 12 orders of magnitude compared to modern supercomputers.
A new molecule discovered by researchers at University of Limerick in Ireland enables fast decision-making in computers, breaking the von Neumann bottleneck. The device can solve problems even if individual components fail, providing smaller, faster, and more energy-efficient computing.
Researchers at George Washington University have created a nanophotonic analog processor capable of solving partial differential equations. The processor can process arbitrary inputs at the speed of light and is integrated at chip-scale.
Researchers at Tokyo Institute of Technology developed a tunable neural network framework that achieves high accuracy and efficiency for sparse CNNs. The new architecture employs a Cartesian-product MAC array and pipelined activation aligners to enable dense computing of sparse convolution, resulting in better resource utilization.