Neural inference at the frontier of energy, space, and time.

Computing, since its inception, has been processor-centric, with memory separated from compute. Inspired by the organic brain and optimized for inorganic silicon, NorthPole is a neural inference architecture that blurs this boundary by eliminating off-chip memory, intertwining compute with memory on-chip, and appearing externally as an active memory chip. NorthPole is a low-precision, massively parallel, densely interconnected, energy-efficient, and spatial computing architecture with a co-optimized, high-utilization programming model. On the ResNet50 benchmark image classification network, relative to a graphics processing unit (GPU) that uses a comparable 12-nanometer technology process, NorthPole achieves a 25 times higher energy metric of frames per second (FPS) per watt, a 5 times higher space metric of FPS per transistor, and a 22 times lower time metric of latency. Similar results are reported for the Yolo-v4 detection network. NorthPole outperforms all prevalent architectures, even those that use more-advanced technology processes.

Convolutional networks for fast, energy-efficient neuromorphic computing.

Deep networks are now able to achieve human-level performance on a broad spectrum of recognition tasks. Independently, neuromorphic computing has now demonstrated unprecedented energy-efficiency through a new chip architecture based on spiking neurons, low precision synapses, and a scalable communication network. Here, we demonstrate that neuromorphic computing, despite its novel architectural primitives, can implement deep convolution networks that (i) approach state-of-the-art classification accuracy across eight standard datasets encompassing vision and speech, (ii) perform inference while preserving the hardware's underlying energy-efficiency and high throughput, running on the aforementioned datasets at between 1,200 and 2,600 frames/s and using between 25 and 275 mW (effectively >6,000 frames/s per Watt), and (iii) can be specified and trained using backpropagation with the same ease-of-use as contemporary deep learning. This approach allows the algorithmic power of deep learning to be merged with the efficiency of neuromorphic processors, bringing the promise of embedded, intelligent, brain-inspired computing one step closer.

Artificial brains. A million spiking-neuron integrated circuit with a scalable communication network and interface.

Inspired by the brain's structure, we have developed an efficient, scalable, and flexible non-von Neumann architecture that leverages contemporary silicon technology. To demonstrate, we built a 5.4-billion-transistor chip with 4096 neurosynaptic cores interconnected via an intrachip network that integrates 1 million programmable spiking neurons and 256 million configurable synapses. Chips can be tiled in two dimensions via an interchip communication interface, seamlessly scaling the architecture to a cortexlike sheet of arbitrary size. The architecture is well suited to many applications that use complex neural networks in real time, for example, multiobject detection and classification. With 400-pixel-by-240-pixel video input at 30 frames per second, the chip consumes 63 milliwatts.

Modeling the GABAergic action of etomidate on the thalamocortical system.

BACKGROUND: We have used a computational model of the thalamocortical system to investigate the effects of a GABAergic anesthetic (etomidate) on cerebral cortical and thalamic neuronal function. We examined the effects of phasic and tonic inhibition, as well as the relative importance of anesthetic action in the thalamus and cortex. METHODS: The amount of phasic GABAergic inhibition was adjusted in the model to simulate etomidate concentrations of between 0.25 and 2 microM, with the concentration range producing unconsciousness assumed to be between 0.25 and 0.5 microM. In addition, we modeled tonic inhibition separately, and then phasic and tonic inhibition together. We also introduced phasic and tonic inhibition into the cerebral cortex and thalamus separately to determine the relative importance of each of these structures to anesthetic-induced depression of the thalamocortical system. RESULTS: Phasic inhibition decreased cortical neuronal firing by 11%-18% in the 0.25-0.5 microM range and by 38% at 2 microM. Tonic inhibition produced similar depression (11%-21%) in the 0.25-0.5 microM range but 65% depression at 2 microM; phasic and tonic inhibition combined produced the most inhibition (76% depression at 2 microM). When the thalamus and cortex were separately subjected to phasic and tonic inhibition, cortical firing rates decreased less compared to when both structures were targeted. In the 0.25-0.5 microM range, cortical firing rate was minimally affected when etomidate action was simulated in the thalamus only. CONCLUSIONS: This computational model of the thalamocortical system indicated that tonic GABAergic inhibition seems to be more important than phasic GABAergic inhibition (especially at larger etomidate concentrations), although both combined had the most effect on cerebral cortical firing rates. Furthermore, etomidate action in the thalamus, by itself, does not likely explain etomidate-induced unconsciousness.