Power/Performance Bits: July 15


Liquefied gas electrolyte Researchers at UC San Diego, U.S. Army Research Laboratory, and South 8 Technologies developed an electrolyte that could enable the replacement of the graphite anode in lithium-ion batteries with lithium-metal. Such a change would increase energy density 50% at the cell level, making for lighter batteries with more capacity. However, lithium-metal anodes are not compa... » read more

Silicon Photonics Begins To Make Inroads


Integrating photons and electrons on the same die is still a long way off, but advances in packaging and improvements in silicon photonics are making it possible to use optical communication for a variety of new applications. Utilizing light-based communication between chips, or in self-contained modules, ultimately could have a big impact on chip design. Photons moving through waveguides ar... » read more

System Bits: July 3


CMU prof gets a shot at new supercomputer The National Energy Research Scientific Computing Center will greet its Perlmutter supercomputing system in early 2020. The Cray-designed machine will be capable of 100 million billion floating operations per second. Zachary Ulissi of Carnegie Mellon University will be among the first researchers to use the supercomputer. "When this machine comes on... » read more

Data Confusion At The Edge


Disparities in pre-processing of data at the edge, coupled with a total lack of standardization, are raising questions about how that data will be prioritized and managed in AI and machine learning systems. Initially, the idea was that 5G would connect edge data to the cloud, where massive server farms would infer patterns from that data and send it back to the edge devices. But there is far... » read more

System Bits: June 10


SlothBot swings through the trees, slowly A robot that doesn’t often move, spending its days, weeks, months, in the forest canopy, monitoring the local environment – that’s SlothBot, from the Georgia Institute of Technology. The robot has two photovoltaic solar panels for its power source. It is designed to stay in the trees for months at a time. It’s gone through trials on the Geor... » read more

System Bits: May 28


Home robotics get cozier Cornell University’s Guy Hoffman was perplexed when he first saw social robots in stores. “I noticed a lot of them had a very similar kind of feature – white and plasticky, designed like consumer electronic devices,” said Hoffman, assistant professor and the Mills Family Faculty Fellow in the Sibley School of Mechanical and Aerospace Engineering. “Especial... » read more

Power/Performance Bits: May 21


More speculative vulnerabilities Security researchers at the Graz University of Technology, KU Leuven, Cyberus Technology, and Worcester Polytechnic Institute point to two new speculative execution vulnerabilities related to Meltdown and Spectre. The first, which they dubbed ZombieLoad, uses a similar approach to Meltdown. After preparing tasks in parallel, the processor needs to discard th... » read more

Power/Performance Bits: May 6


Compressing objects Computer scientists at MIT propose a way to improve data compression in memory by focusing on objects rather than cache lines. "The motivation was trying to come up with a new memory hierarchy that could do object-based compression, instead of cache-line compression, because that's how most modern programming languages manage data," said Po-An Tsai, a graduate student at... » read more

System Bits: April 23


AI tool can clean up dirty data Researchers at the University of Waterloo, collaborating with colleagues at the University of Wisconsin and Stanford University, came up with HoloClean, an artificial intelligence tool to comb through dirty data and to detect information errors. “More and more machines are making decisions for us, so all our lives are touched by dirty data daily,” said Ih... » read more

Power/Performance Bits: April 16


Faster CNN training Researchers at North Carolina State University developed a technique that reduces training time for deep learning networks by more than 60% without sacrificing accuracy. Convolutional neural networks (CNN) divide images into blocks, which are then run through a series of computational filters. In training, this needs to be repeated for the thousands to millions of images... » read more

← Older posts