System Bits: Nov. 7


Exposing logic errors in deep neural networks In a new approach meant to brings transparency to self-driving cars and other self-taught systems, researchers at Columbia and Lehigh universities have come up with a way to automatically error-check the thousands to millions of neurons in a deep learning neural network. Their tool — DeepXplore — feeds confusing, real-world inputs into the ... » read more

System Bits: Oct. 31


Software enables cars to auto-report diagnostics Thanks to researchers at MIT, it may soon be possible to hop into a ride-share car, glance at a smartphone app, and tell the driver that the car’s left front tire needs air, its air filter should be replaced next week, and its engine needs two new spark plugs. [caption id="attachment_409967" align="alignnone" width="300"] A new smartphone a... » read more

Power/Performance Bits: Oct. 31


Battery material supplies Researchers at MIT, the University of California at Berkeley, and the Rochester Institute of Technology conducted an analysis of whether there are enough raw materials to support increased lithium-ion battery production, expected to grow significantly due to electric vehicles and grid-connected battery systems. They conclude that while in the near future there shou... » read more

Making high-capacity data caches more efficient


Source: Researchers from MIT, Intel, and ETH Zurich Xiangyao Yu (MIT), Christopher J. Hughes (Intel), Nadathur Satish (Intel) Onur Mutlu (ETH Zurich), Srinivas Devadas (MIT) Technical Paper link MIT News article As the transistor counts in processors have gone up, the relatively slow connection between the processor and main memory has become the chief impediment to improving comp... » read more

System Bits: Oct. 24


Optical communication on silicon chips With the huge increase in computing performance in recent decades achieved by squeezing ever more transistors into a tighter space on microchips, at the same time this downsizing has also meant packing the wiring within microprocessors ever more tightly together. This has led to effects such as signal leakage between components, which can slow down commun... » read more

Power/Performance Bits: Oct. 24


Molecular storage Chemists at the Institut Charles Sadron and Aix-Marseille University used mass spectrometry to read several bytes of data recorded on the molecular scale with synthetic polymers, setting a new benchmark for the amount of data stored as a sequence of molecular units (monomers) that can be read. Polymers have great potential since, to record a bit, their component monomers r... » read more

How Neural Networks Think (MIT)


Source: MIT’s Computer Science and Artificial Intelligence Laboratory, David Alvarez-Melis and Tommi S. Jaakkola Technical paper link MIT article General-purpose neural net training Artificial-intelligence research has been transformed by machine-learning systems called neural networks, which learn how to perform tasks by analyzing huge volumes of training data, reminded MIT research... » read more

System Bits: Oct. 17


Piezoelectric, ingestible sensors With an aim to help doctors diagnose gastrointestinal disorders that slow down the passage of food through the digestive tract, MIT and Brigham and Women’s Hospital researchers have built a flexible sensor that can be rolled up and swallowed. Once ingested, the sensor adheres to the stomach wall or intestinal lining, where it can measure the rhythmic con... » read more

System Bits: Oct. 10


Fast-moving magnetic particles for data storage According to MIT researchers, an exotic kind of magnetic behavior discovered just a few years ago holds great promise as a way of storing data — one that could overcome fundamental limits that might otherwise be signaling the end of Moore’s Law. Rather than reading and writing data one bit at a time by changing the orientation of magnetize... » read more

System Bits: Oct. 3


Polariton graphs In a development that a team of researchers from the UK and Russia say could eventually surpass the capabilities of even the most powerful supercomputers, a type of ‘magic dust’ — which combines light and matter — can be used to solve complex problems. Hailing from the University of Cambridge, University of Southampton and Cardiff University in the UK and the Skolk... » read more

← Older posts