Lifeboat Foundation

Scientists create a brain-like organic transistor that can learn and evolve.

A group of researchers at Sandia National Laboratories have developed a tool that can cross-train standard convolutional neural networks (CNN) to a spiking neural model that can be used on neuromorphic processors. The researchers claim that the conversion will enable deep learning applications to take advantage of the much better energy efficiency of neuromorphic hardware, which are designed to mimic the way the biological neurons work.

The tool, known as Whetstone, works by adjusting artificial neuron behavior during the training phase to only activate when it reaches an appropriate threshold. As a result, neuron activation become a binary choice – either it spikes or it doesn’t. By doing so, Whetstone converts an artificial neural network into a spiking neural network. The tool does this by using an incremental “sharpening process” (hence Whetstone) through each network layer until the activation becomes discrete.

According to Whetstone researcher Brad Aimone, this discrete activation greatly minimizes communication costs between the layers, and thus energy consumption, but with only minimal loss of accuracy. “We continue to be impressed that without dramatically changing what the networks look like, we can get very close to a standard neural net [in accuracy],” he says. “We’re usually within a percent or so on performance.”

Continue reading “One Step Closer to Deep Learning on Neuromorphic Hardware” »

Neuromorphic systems carry out robust and efficient neural computation using hardware implementations that operate in physical time. Typically they are event- or data-driven, they employ low-power, massively parallel hybrid analog/digital VLSI circuits, and they operate using the same physics of computation used by the nervous system. Although there are several forums for presenting research achievements in neuromorphic engineering, none are exclusively dedicated to this increasingly large research community. Either because they are dedicated to single disciplines, such as electrical engineering or computer science, or because they serve research communities which focus on analogous areas (such as biomedical engineering or computational neuroscience), but with fundamentally different goals and objectives. The mission of Neuromorphic Engineering is to provide a publication medium dedicated exclusively and specifically to this field. Topics covered by this publication include:  Analog and hybrid analog/digital electronic circuits for implementing neural processes, such as conductances, neurons, synapses, plasticity mechanisms, photoreceptors, cochleae, etc.  Neuromorphic circuits and systems for implementing real-time event-based neural processing architectures.  Hardware models of neural and sensorimotor processing systems, such as selective attention systems, coordinate transformation systems, auditory and/or visual processing systems, sensory fusion systems, etc.  Implementations of neural computational systems found in insects, birds, mammals, etc.  Embedded neuromorphic systems, including actuated or robotic platforms which process sensory signals and interact with the environment using event-based sensors and circuits. To ensure high quality and state-of-the-art material, publications should demonstrate experimental results, using physical implementations of neuromorphic systems, and possibly show the links between the artificial system and the neural/biological one they model.

Discover how and why neuromorphic and probabilistic computing solutions represent the next wave of AI capabilities.

Artificial life (often abbreviated ALife or A-Life) is a field of study wherein researchers examine systems related to natural life, its processes, and its evolution, through the use of simulations with computer models, robotics, and biochemistry.

A new photonic chip design drastically reduces energy needed to compute with light, with simulations suggesting it could run optical neural networks 10 million times more efficiently than its electrical counterparts.

Image: courtesy of the researchers, edited by MIT News.

The most essential Pentagon suppliers will be the ones that master robotics and artificial intelligence.

Look across the Potomac River toward Rosslyn, where the corporate logos of government contractors crown a parade of office towers that follows the river past the Pentagon. The skyline, like America’s defense industrial landscape, is changing. Soon, 25,000 Amazon employees will be climbing the Metro escalators to work in Crystal City each morning along with the tens of thousands of workers from military, intelligence, and the defense industry organizations.

The arrival of Amazon’s HQ2 in the cradle of U.S. government contracting comes at a portentous time for the Defense Department. Technology is altering what makes us strong, prosperous, and secure. The defense industrial base is becoming the strategic innovation base. Today’s leading digital companies have disrupted every industry they have touched, from publishing to automotive. Could Amazon and the rest of the “FAANG companies”—Facebook, Apple, Netflix, and Google—or one of a handful of pure-play artificial-intelligence companies, such as the authors’ SparkCognition, become fixtures of this new industrial base?

Continue reading “A New Industrial Base Is Taking Shape. Call It the Military-AI Complex” »

In 2017, an online disinformation campaign spread against the “White Helmets,” claiming that the group of aid volunteers was serving as an arm of Western governments to sow unrest in Syria.

A new photonic chip could run optical neural networks 10 million times more efficiently than conventional chips.

The classical physical limit for computing energy is the Landauer limit that sets a lower bound to the minimum heat dissipated per bit erasing operation. Performance below the thermodynamic (Landauer) limit for digital irreversible computation is theoretically possible in this device. The proposed accelerator can implement both fully connected and convolutional networks.

Previous photonic chips had bulky optical components that limited their use to relatively small neural networks. MIT researchers have a new photonic accelerator that uses more compact optical components and optical signal-processing techniques, to drastically reduce both power consumption and chip area. That allows the chip to scale to neural networks several orders of magnitude larger than its counterparts.

Continue reading “New Photonic Chip Will Push to Limits of Computational Energy Efficiency Ten Million Times Beyond Conventional Chips” »