Larry Hardesty, MIT News Office
Articles
Integrating optical components into existing chip designs
Researchers at MIT, the University of California at Berkeley, and Boston University have developed a technique that would allow an addition of optical communication components to existing chips with little design modification.
Neural network chip reduces power consumption
MIT researchers have developed a chip designed to reduce neural networks’ power consumption by up to 95%, making them practical for battery-powered devices.
Energy efficient encryption for the IoT
Special-purpose chip developed by MIT researchers is designed to reduce power consumption of public-key encryption by 99.75% while increasing speed 500-fold.
Computer systems predict objects’ responses to physical forces
MIT researchers believe they can help answer questions about what information-processing resources human beings use at what stages of development by building computer systems that approximate these capacities, which might generate some insights useful for robotic vision systems.
Processing a neural network’s mind and its ability to process language
MIT researchers have developed a technique illuminates the inner workings of artificial intelligence systems that process language, which could improve overall efficiency for machines.
Device makes power conversion more efficient
Researchers at MIT, IQE, Columbia University, IBM, and the Singapore-MIT Alliance for Research and Technology have created a device that enables gallium nitride power devices to handle voltages of 1,200 volts, which is double their current ability, to help cut energy waste in electric vehicles, data centers, and the power grid.
Drones relay RFID signals for inventory control
MIT researchers have developed a system that enables aerial drones to read radio frequency ID (RFID) tags from tens of meters away while identifying the tags’ locations with a small error window. The system is designed for use used in large warehouses for both continuous monitoring to prevent inventory mismatches and the ability to locate individual items.
Automatic image retouching for phone, vision applications
Researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and Google have developed a system can apply a range of styles in real-time, so that the viewfinder displays the enhanced image, which could be beneficial for machine vision applications.
Bringing neural networks to cellphones
MIT researchers have designed new methods for paring down neural networks so that they’ll run more efficiently on handheld devices.
Algorithms for wearable devices help users, robots avoid obstacles
Scientists at MIT have developed algorithms that help power a prototype system for helping visually impaired users avoid obstacles and identify objects and can also be used as sensors for robots.