Break it down: A new way to address common computing problem

Science Daily  August 5, 2020 The volume of data, inadequate computational resources to handle an oversized problem, security and privacy concerns are some of the challenges in solving large-scale linear inverse problems (LIP). Researchers at Washington University have developed the parallel residual projection (PRP), a parallel computational framework involving the decomposition of a large-scale LIP into sub-problems of low complexity and the fusion of the sub-problem solutions to form the solution to the original LIP. They analyzed the convergence properties of the PRP and accentuate its benefits through its application to complex problems of network inference and gravimetric survey. They […]

Computing with spins of light

Phys.org  January 21, 2020 In the XY model of a spin system used for solving complex problems particles form vortex-like patterns around multiple focal points. An international team of researchers (Japan, USA -Stanford University) improved an experimental set up that emulates the XY model by designing a laser system that generates pulses of light particles within a one-kilometer-long optical fiber cavity. The pulses generated 5,000 spins, i.e. flows of light particles going down 5,000 different drains, within the cavity. It enabled long simulations over several minutes within a stable system. This improves on previous oscillators that produced only 100 spins […]

Reorganizing a computer chip: Transistors can now both process and store information

Science Daily  December 9, 2019 Researchers at Purdue University used a semiconductor that has ferroelectric properties. This way two materials become one material so that the interface issues are not a problem. The result is a so-called ferroelectric semiconductor field-effect transistor, built in the same way as transistors currently used on computer chips. The material, alpha indium selenide has ferroelectric properties and a much smaller band gap, making it possible for the material to be a semiconductor without losing ferroelectric properties. The researchers have created a more feasible way to combine transistors and memory on a chip, potentially bringing faster […]

DARPA uses nature as a muse for new computing model

Defense Systems  August 9, 2019 Taking a cue from nature’s efficiency and precision, DARPA is looking for concepts that exploit the interplay between complex dynamical behaviors and intrinsic properties of materials to develop novel computing models for the purpose of tackling current hard computation problems, according to an Aug. 1 solicitation. The two-phase, 18-month program has two technical areas: theory and design, then application development…read more.

Virtually energy-free superfast computing invented by scientists using light pulses

Science Daily  May 15, 2019 An international team of researchers (Germany, USA- UC Santa Barbara, Russia, the Netherlands) utilized the efficient interaction mechanism of coupling between spins and terahertz electric field, which was discovered by the same team. They developed and fabricated a very small antenna on top of the magnet to concentrate and thereby enhance the electric field of light. This strongest local electric field was sufficient to navigate the magnetization of the magnet to its new orientation in just one trillionth of a second without increasing the temperature. Future storage devices would also exploit the excellent spatial definition […]

The cost of computation

Science Daily  April 8, 2019 An international team of researchers (USA – Arizona State University, Santa Fe Institute, Austria) reviews some of the recent work on the ‘stochastic thermodynamics of computation. After reviewing the salient parts of information theory, computer science theory, and stochastic thermodynamics, they summarize what has been learned about the entropic costs of performing a broad range of computations, extending from bit erasure to loop-free circuits to logically reversible circuits to information ratchets to Turing machines. These results reveal new, challenging engineering problems for how to design computers to have minimal thermodynamic costs. They also allow us […]

Engineers demonstrate metamaterials that can solve equations

Science Daily  March 22, 2019 Signal processing of light waves can be used to represent certain mathematical functions and to perform computational tasks on signals or images in an analog fashion using complex systems of bulk optical elements such as lenses, filters, and mirrors. Researchers at the University of Pennsylvania have demonstrated that specially designed nanophotonic structures can take input waveforms encoded as complex mathematical functions, manipulate them, and provide an output that is the integral of the functions. The results, demonstrated for microwaves, provide a route to develop chip-based analog optical computers and computing elements… read more. TECHNICAL ARTICLE 

Computational intelligence-inspired clustering in multi-access vehicular networks

Phys.org June 28, 2018 Vehicular ad hoc networks (VANETs) have attracted great interest for improving communications between vehicles using infrastructure-less wireless technologies. To overcome the shortcomings of the VANETs, an international team of researchers (Japan, Finland) has proposed a two-level clustering approach where cluster head nodes in the first level try to reduce the MAC layer contentions for vehicle-to-vehicle (V2V) communications, and cluster head nodes in the second level are responsible for providing a gateway functionality between V2V and LTE. Simulation results under various Network conditions show that the proposed protocol can achieve 23% throughput improvement in high-density scenarios compared […]

‘Breakthrough’ algorithm exponentially faster than any previous one

Eurekalert June 28, 2018 Traditionally algorithms for optimization problems narrow down the search space for the best solution one step at a time. In contrast, the new algorithm developed by researchers at Harvard University samples a variety of directions in parallel. Based on that sample, the algorithm discards low-value directions from its search space and chooses the most valuable directions to progress towards a solution. Using a data set of two million taxi trips from the New York City taxi and limousine commission, the adaptive-sampling algorithm found solutions 6 times faster. One of the biggest challenges in machine learning is […]