Welcome to the August 6, 2018 edition of ACM TechNews, providing timely information for IT professionals three times a week.
ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version.
The online version now has a button at the top labeled "Show Headlines."
|
|
|
NIST Chip Lights Up Optical Neural Network Demo
National Institute of Standards and Technology
Laura Ost
July 26, 2018
U.S. National Institute of Standards and Technology (NIST) researchers say they have created a neural network-equipped silicon chip that distributes optical signals precisely across a miniature grid. The team says light's benefits include the elimination of interference caused by electrical charge, as well as extending transmission distance and speed. NIST's grid-on-a-chip stacks two layers of photonic silicon nitride waveguides in a three-dimensional (3D) architecture to facilitate complex routing schemes essential to the mimicry of neural systems. The 3D grid has 10 inputs that each link to 10 outputs, for a total of 100 receivers. Accompanying software can automatically generate signal routing, with adjustable levels of connectivity between neurons. Said NIST’s Jeff Chiles, “Light’s advantages could improve the performance of neural nets for scientific data analysis such as searches for Earth-like planets and quantum information science, and accelerate the development of highly intuitive control systems for autonomous vehicles.”
|
Memory-Processing Unit (MPU) Could Bring Memristors to the Masses
The Michigan Engineer News Center
Kate McAlpine
July 26, 2018
The University of Michigan's Wei Lu and his colleagues have developed a method for assembling memristors on a chip for use in general computing while reducing their energy consumption 100-fold. Key to this breakthrough is digitizing current outputs and plotting out large mathematical problems into smaller blocks within the array, which the researchers call memory-processing units (MPUs). Wu says MPUs could be especially helpful for deploying machine learning and artificial intelligence algorithms, as well as for executing tasks based on matrix operations, such as weather prediction models. The least complex mathematical matrices can map directly onto the grid of memristors. After the memristors are programmed to represent the numbers, operations that multiply and add the matrices are performed concurrently. Lu says, "We don't need to manually multiply and sum in a processor."
|
UCF Professor Discovers First-of-Its-Kind Material for the Quantum Age
University of Central Florida
Allison Hurtado
August 1, 2018
A University of Central Florida (UCF) researcher says he has discovered a material that could be used as a building block for quantum technology. The material, Hf2Te2P, is chemically composed of hafnium, tellurium, and phosphorus, and is the first material that has multiple quantum properties. UCF's Madhab Neupane says, "Our discovery takes us one step closer to the application of quantum materials and helps us gain a deeper understanding of the interactions between various quantum phases." The material has more than one electron pattern that develops within its electronic structure, giving it a range of quantum properties. Neupane says this material will boost computing power for big data in new devices, and greatly lower the amount of energy required to power electronics.
|
Greek Mathematician Constantinos Daskalakis Wins Prestigious Award
Greek Reporter
Philip Chrysopoulos
August 2, 2018
The International Mathematical Society has named Massachusetts Institute of Technology professor Constantinos Daskalakis to receive this year’s Rolf Nevanlinna Prize. The society says it singled out Daskalakis because his work has "shaped our comprehension of computational complexity to fundamental problems" related to math and the economy. The Greek scientist was awarded ACM's 2008 Doctoral Dissertation Award for advancing the understanding of behavior in complex networks of interacting individuals. Daskalakis' dissertation provided a novel algorithmic view on game theory and the concept of the Nash equilibrium, and it also earned him a spot as co-recipient, along with Christos Papadimitriou and Paul W. Goldberg, of the 2008 Kalai Prize for outstanding articles at the interface of computer science and game theory. The Rolf Nevanlinna Prize is given every four years to researchers younger than 40 who have contributed to the advancement of the mathematical aspects of computer science.
|
U of T Launches Canada's First Engineering Undergraduate Program in Machine Intelligence
U of T News
Tyler Irving
July 25, 2018
The University of Toronto (U of T) has unveiled Canada's first undergraduate engineering science program in machine intelligence, scheduled to launch in September with more than 40 students enrolled so far. Says U of T's Deepa Kundur, "U of T Engineering has long been an international leader in machine intelligence, making this the perfect place to launch this pioneering program." In January, the university launched the Master of Engineering in Analytics program, which provides instruction in techniques and strategies to translate large data sets into useful insights. The university’s new Myhal Center for Engineering Innovation & Entrepreneurship will host the new Institute for Robotics and Mechatronics, as well as prototyping facilities, and space for interaction among students, faculty, alumni, industry partners, and business mentors.
|
The 2018 Top Programming Languages
IEEE Spectrum
Stephen Cass
July 31, 2018
IEEE Spectrum's fifth annual interactive ranking of the top programming languages draws on metrics from multiple sources to rank 47 languages. Because different programmers have different needs and domains of interest, the IEEE Spectrum team chose to develop an interactive application that enables the user to choose how the metrics are weighted when combined. The ranking includes a default weighting, as well as other presets that focus on different aspects of coding. IEEE’s ranking for its typical members found Python as the top programming language with a score of 100, followed by C++ with 98.4 and C with 98.2. Python continues to gain popularity among programmers because it is now an embedded language; many modern microcontrollers now have more than enough power to host a Python interpreter, making the language useful in specific applications to play with attached hardware via an interactive prompt or dynamically reload scripts in real time.
|
Lip-Reading AI Could Help the Deaf—or Spies
Science
Matthew Hutson
July 31, 2018
An artificial intelligence (AI) program from DeepMind can read lips better than professional lip readers, after reviewing thousands of hours of YouTube videos along with transcripts via machine learning. The researchers tested the program on 37 minutes of video it had not previously viewed, and it misidentified only 41% of the words. In comparison, the best previous computer method, which focuses on individual letters instead of phonemes, had a 77% word error rate, while professional lip readers had a 93% error rate in the same test, which lacked context or body language. Columbia University's Hassan Akbari says the AI, if incorporated into a phone, would enable hearing-impaired users to have a "translator" with them wherever they go. Helen Bear at Queen Mary University of London in the U.K. envisions applications for the program that include analyzing security video, interpreting historical footage, or understanding what a Skype partner is saying when the audio goes dead.
|
BMI System Lets Users Control Robotic Arm While Their Hands Are Busy
Tech Xplore
Bob Yirka
July 26, 2018
Researchers at the Advanced Telecommunications Research Institute International in Japan have created a brain-machine interface (BMI) for manipulating a robotic arm without requiring the use of hands. Their system enables the user to control three limbs at once, including two natural arms and the robotic arm. The researchers demonstrated the system by having volunteers balance a ball on a flat surface while the system recorded their brain wave activity; afterward, when a volunteer would think about balancing the ball, the system would recognize the brain wave pattern and move the robot arm to move the ball in a similar way. The researchers found some volunteers were more successful than others in having the robot arm balance the ball, which the researchers said probably was due to the volunteers, not the system.
|
SDSC's Comet Supercomputer Extended Into 2021
HPCwire
August 1, 2018
The San Diego Supercomputer Center (SDSC) at the University of California, San Diego has received a supplemental grant from the U.S. National Science Foundation (NSF) of nearly $2.4 million to extend operations of its Comet supercomputer through March 2021. The extension brings the value of the total Comet program to more than $27 million. Comet is capable of an overall peak performance of two petaflops. As a direct response to the growing demand for graphics-processing units (GPUs) among a wide range of research domains, SDSC recently doubled the number of GPUs on Comet. The NSF award makes Comet the largest GPU resource available through the eXtreme Science and Engineering Discovery Environment (XSEDE) program.
|
UTA Technology Could Change Way Computers Dissipate Heat
University of Texas at Arlington
Herb Booth
July 31, 2018
Researchers at the University of Texas at Arlington (UTA) have developed a novel cold electron transistor that drastically reduces the amount of energy it requires to operate. They say the technology could reduce power dissipation on transistors inside consumer electronics by a factor of 100. In addition, the transistors could yield huge energy savings for organizations that need large amounts of transistor operations in their data centers. UTA's Seong Jin Koh says the new technology could result in computers or cellphones that only need to be charged every few weeks. The researchers see commercial and military applications for the technology, which would reduce the size of batteries needed to run electronic equipment.
|
3D Printing the Next Generation of Batteries
Carnegie Mellon University
Emily Durham
July 30, 2018
Carnegie Mellon University and Missouri University of Science and Technology researchers have developed a method of three-dimensionally (3D) printing battery electrodes that creates a 3D microlattice structure with controlled porosity. They say the technique vastly improves the capacity and charge-discharge rates for lithium-ion batteries. The method marks a major advance in the printing of complex geometries for 3D battery architectures, as well as a step toward geometrically optimizing 3D configurations for electrochemical energy storage. The microlattice structure used as lithium-ion batteries' electrodes provides a fourfold increase in battery capacity, and a twofold increase in areal capacity when compared to a solid block electrode. In addition, the electrodes retain their complex 3D lattice structures after 40 electrochemical cycles, demonstrating their high energy capacity for the same weight, or the same capacity at a vastly reduced weight.
|
Trapping Light That Doesn't Bounce Off Track for Faster Electronics
Purdue University News
Kayla Wiles
July 30, 2018
An international project led by Purdue University researchers has yielded a metamaterial "cladding" to block the leakage of light from computer chips, in a step toward faster electronics. Purdue's Zubin Jacob says, "We want the bits of information that we are sending in the waveguide to travel along tight bends and simultaneously not be lost as heat." The team was able to prevent light from leaking off track into other waveguides and causing information-mixing or "crosstalk" by controlling the cladding's anisotropy to enable light to travel at different velocities in different directions; as a result, light-carried bits stay strongly contained in the waveguide. Purdue's Saman Jahani says the waveguide has extreme skin depth to ensure leakage is minuscule, which "can pave the way for dense photonic integration on a computer chip without worrying about light leakage."
|
|
