Brain-inspired computing; direct laser writing; photonics breakthrough.
Lawrence Livermore National Laboratory (LLNL) has purchased a brain-inspired supercomputing platform for deep learning developed by IBM Research.
Based on a neurosynaptic computer chip called IBM TrueNorth, the scalable platform will process the equivalent of 16 million neurons and 4 billion synapses. It will consume the energy equivalent of a tablet computer.
Under the terms of the contract, LLNL will receive a 16-chip TrueNorth system. A single TrueNorth processor consists of 5.4 billion transistors wired together to create an array of 1 million digital neurons that communicate with one another via 256 million electrical synapses. At 0.8 volts, it consumes 70 milliwatts of power running in real time and delivers 46 giga synaptic operations per second. The chip is fabricated based on a 28nm LPP process from Samsung.
The technology represents a departure from computer design that has been prevalent for the past 70 years. Like the human brain, neurosynaptic or neuromorphic systems require less electrical power and volume. It is able to perform at exascale speeds, 50 times (or two orders of magnitude) faster than today’s most advanced petaflop (quadrillion floating point operations per second) systems.
Neuromorphic technology, sometimes called brain-inspired computing, is a paradigm shift that breaks away from Moore’s Law. Neuromorphic chips don’t require costly leading-edge processes.
The brain-like, neural network design of the IBM Neuromorphic System is able to infer complex cognitive tasks such as pattern recognition and integrated sensory processing far more efficiently than conventional chips.
The new system will be used to explore new computing capabilities important to the National Nuclear Security Administration’s (NNSA) missions in cyber security. NNSA’s Advanced Simulation and Computing (ASC) program will evaluate machine learning applications, deep learning algorithms and architectures and conduct general computing feasibility studies.
“Neuromorphic computing opens very exciting new possibilities and is consistent with what we see as the future of the high performance computing and simulation at the heart of our national security missions,” said Jim Brase, LLNL deputy associate director for Data Science, in a statement. “The potential capabilities neuromorphic computing represents and the machine intelligence that these will enable will change how we do science.”
“The low power consumption of these brain-inspired processors reflects the industry’s desire and a creative approach to reducing power consumption in all components for future systems as we set our sights on exascale computing,” said Michel McCoy, LLNL program director for Weapon Simulation and Computing.
“The delivery of this advanced computing platform represents a major milestone as we enter the next era of cognitive computing,” said Dharmendra S. Modha, an IBM Fellow, chief scientist and brain-inspired computing at IBM Research – Almaden. “This collaboration will push the boundaries of brain-inspired computing to enable future systems that deliver unprecedented capability and throughput, while helping to minimize the capital, operating and programming costs – keeping our nation at the leading edge of science and technology.”
TrueNorth was originally developed under the auspices of Defense Advanced Research Projects Agency’s (DARPA) Systems of Neuromorphic Adaptive Plastic Scalable Electronics (SyNAPSE) program in collaboration with Cornell University.
Direct laser writing
The University of Missouri has developed a one-step direct laser writing (DLW) technology for use in synthesizing and patterning novel materials.
Researchers used the technique to pattern hybrid materials based on molybdenum disulfide (MoS2) and carbon. This, in turn, enabled the development of hydrogen evolution reaction catalysts.
Using computer-controlled laser beams, the catalysts could be used to fabricate energy storage units, such as microbatteries, micro fuel cells and other products.
DLW could represent a major breakthrough. MoS2/carbon hybrid materials are promising electrocatalysts, but a method to synthesize and pattern these materials is a major challenge, according to researchers. With DLW, however, researchers synthesized small-sized MoS2 nanoparticles, which were anchored to a carbon matrix. The hybrid materials exhibit good catalytic performance and stability, according to researchers.
“The direct laser writing (DLW) method and technique has seen a rapid advancement in the past decade,” said Jian Lin, an assistant professor at the University of Missouri, in a statement. “The main goal of our research was to find an efficient and cost-effective way to integrate nanostructures with micro energy storage units for applications in micro-electronics. Our lab decided to test whether catalysts could be synthesized and patterned on any surface by a one-step laser processing method to produce microbatteries and micro fuel cells in the shapes dictated by computer programs.
“This is the first step in manufacturing micro fuel cells that convert chemical energy into electrical energy and batteries that can integrate into microcircuits,” said Lin. “Also this technique has been proven to produce microsupercapacitors. By honing the process, handheld device and smartphone manufacturers will be able to produce components in whatever shape or size they choose, greatly impacting the size of these devices. Also, manufacturers will be able to choose more environmentally friendly catalysts for generating energy such as hydrogen or oxygen, which are considered cleaner fuels. The possibilities will be endless.”
IRT Nanoelec, an R&D consortium headed by CEA-Leti, has made a major breakthrough in silicon photonics.
The consortium has demonstrated the first co-integration of two components–a III-V/silicon laser and a silicon Mach Zehnder modulator. This combination, in turn, enabled 25-Gbps transmissions on a single channel. Typically, this transmission is achieved using an external source over a 10 km single-mode fiber.
To achieve these results, researchers integrated silicon photonics circuits with a modulator. Initially, the devices were processed on a 200mm silicon-on-insulator (SOI) wafer. Then, a two-inch wafer of III-V material was directly bonded on the wafer. The hybrid wafer was processed using conventional semiconductor and/or MEMS process steps to produce an integrated modulator-and-laser transmitter.
“IRT Nanoelec and its partners on this project, Leti, STMicroelectronics, Samtec and Mentor Graphics, are paving the way to integrating this technology in next-generation transceivers for optical data links,” said Stéphane Bernabé, project manager at IRT Nanoelec.
IRT Nanoelec launched its silicon photonics program in 2012, with core members Mentor Graphics, STMicroelectronics and CNRS. The program brings together, under one roof, the expertise and equipment needed to address the entire photonics-on-silicon value chain.