Loihi is a 60-mm2 chip fabricated in Intels 14-nm process that advances the state-of-the-art modeling of spiking neural networks in silicon. It integrates a wide range of novel features for the field, such as hierarchical connectivity, dendritic compartments, synaptic delays, and, most importantly, programmable synaptic learning rules. Running a spiking convolutional form of the Locally Competitive Algorithm, Loihi can solve LASSO optimization problems with over three orders of magnitude superior energy-delay-product compared to conventional solvers running on a CPU iso-process/voltage/area. This provides an unambiguous example of spike-based computation, outperforming all known conventional solutions.

Loihi: A Neuromorphic Manycore Processor with On-Chip Learning

Processing-in-memory (PIM) is a promising solution to address the "memory wall" challenges for future computer systems. Prior proposed PIM architectures put additional computation logic in or near memory. The emerging metal-oxide resistive random access memory (ReRAM) has showed its potential to be used for main memory. Moreover, with its crossbar array structure, ReRAM can perform matrix-vector multiplication efficiently, and has been widely studied to accelerate neural network (NN) applications. In this work, we propose a novel PIM architecture, called PRIME, to accelerate NN applications in ReRAM based main memory. In PRIME, a portion of ReRAM crossbar arrays can be configured as accelerators for NN applications or as normal memory for a larger memory space. We provide microarchitecture and circuit designs to enable the morphable functions with an insignificant area overhead. We also design a software/hardware interface for software developers to implement various NNs on PRIME. Benefiting from both the PIM architecture and the efficiency of using ReRAM for NN computation, PRIME distinguishes itself from prior work on NN acceleration, with significant performance improvement and energy saving. Our experimental results show that, compared with a state-of-the-art neural processing unit design, PRIME improves the performance by ~2360× and the energy consumption by ~895×, across the evaluated machine learning benchmarks.

PRIME: a novel processing-in-memory architecture for neural network computation in ReRAM-based main memory

In this paper, the recent progress of synaptic electronics is reviewed. The basics of biological synaptic plasticity and learning are described. The material properties and electrical switching characteristics of a variety of synaptic devices are discussed, with a focus on the use of synaptic devices for neuromorphic or brain-inspired computing. Performance metrics desirable for large-scale implementations of synaptic devices are illustrated. A review of recent work on targeted computing applications with synaptic devices is presented.

https://iopscience.iop.org/article/10.1088/0957-4484/24/38/382001/pdf

Synaptic electronics: materials, devices and applications

Guided by brain-like ‘spiking’ computational frameworks, neuromorphic computing—brain-inspired computing for machine intelligence—promises to realize artificial intelligence while reducing the energy requirements of computing platforms. This interdisciplinary field began with the implementation of silicon circuits for biological neural routines, but has evolved to encompass the hardware implementation of algorithms with spike-based encoding and event-driven representations. Here we provide an overview of the developments in neuromorphic computing for both algorithms and hardware and highlight the fundamentals of learning and hardware frameworks. We discuss the main challenges and the future prospects of neuromorphic computing, with emphasis on algorithm–hardware codesign. The authors review the advantages and future prospects of neuromorphic computing, a multidisciplinary engineering concept for energy-efficient artificial intelligence with brain-inspired functionality.

Towards spike-based machine intelligence with neuromorphic computing.

Dense crossbar arrays of non-volatile memory (NVM) devices represent one possible path for implementing massively-parallel and highly energy-efficient neuromorphic computing systems. We first revie...

https://www.tandfonline.com/doi/pdf/10.1080/23746149.2016.1259585

Neuromorphic computing using non-volatile memory

SRAM cell stability will be a primary concern for future technologies due to variability and decreasing power supply voltages. 6T-SRAM can be optimized for stability by choosing the cell layout, device threshold voltages, and the /spl beta/ ratio. 8T-SRAM, however, provides a much greater enhancement in stability by eliminating cell disturbs during a read access, thus facilitating continued technology scaling. We demonstrate the smallest 6T (0.124/spl mu/m/sup 2/ half-cell) and full 8T (0.1998/spl mu/m/sup 2/) cells to date.

https://www.bioee.ee.columbia.edu/courses/upload/Bibliography/chang_vlsitech_2005.pdf?origin=publication_detail

Stable SRAM cell design for the 32 nm node and beyond

An eight-transistor (8T) cell is proposed to improve variability tolerance and low-voltage operation in high-speed SRAM caches. While the cell itself can be designed for exceptional stability and write margins, array-level implications must also be considered to achieve a viable memory solution. These constraints can be addressed by modifying traditional 6T-SRAM techniques and conceding some design complexity and area penalties. Altogether, 8T-SRAM can be designed without significant area penalty over 6T-SRAM while providing substantially improved variability tolerance and low-voltage operation with no need for secondary or dynamic power supplies. The proposed 8T solution is demonstrated in a high-performance 32 kb subarray designed in 65 nm PD-SOI CMOS that operates at 5.3 GHz at 1.2 V and 295 MHz at 0.41 V.

An 8T-SRAM for Variability Tolerance and Low-Voltage Operation in High-Performance Caches

Efforts to achieve the long-standing dream of realizing scalable learning algorithms for networks of spiking neurons in silicon have been hampered by (a) the limited scalability of analog neuron circuits; (b) the enormous area overhead of learning circuits, which grows with the number of synapses; and (c) the need to implement all inter-neuron communication via off-chip address-events. In this work, a new architecture is proposed to overcome these challenges by combining innovations in computation, memory, and communication, respectively, to leverage (a) robust digital neuron circuits; (b) novel transposable SRAM arrays that share learning circuits, which grow only with the number of neurons; and (c) crossbar fan-out for efficient on-chip inter-neuron communication. Through tight integration of memory (synapses) and computation (neurons), a highly configurable chip comprising 256 neurons and 64K binary synapses with on-chip learning based on spike-timing dependent plasticity is demonstrated in 45nm SOI-CMOS. Near-threshold, event-driven operation at 0.53V is demonstrated to maximize power efficiency for real-time pattern classification, recognition, and associative memory tasks. Future scalable systems built from the foundation provided by this work will open up possibilities for ubiquitous ultra-dense, ultra-low power brain-like cognitive computers.

A 45nm CMOS neuromorphic chip with a scalable architecture for learning in networks of spiking neurons

A switched-capacitor DC-DC voltage converter in 45nm SOI CMOS leverages on-chip trench capacitors to achieve 90% efficiency at an output of 2.3A/mm2 for 2V-to-0.95V conversion at 100MHz. Operation in step-up and step-down modes is demonstrated. Combined with stacked voltage domains, self-regulation capability enables further efficiency improvement.

A fully-integrated switched-capacitor 2∶1 voltage converter with regulation capability and 90% efficiency at 2.3A/mm 2

A multi-chip module is disclosed in which a first die connects to a second set of die via a set of C4 connections within a single package. Low resistivity signal posts are provided within the lateral separation between adjacent die in the second set of die. These signal posts are connectable to externally supplied power signals. The power signals provided to the signals posts are routed to circuits within the second set of die over relatively short metallization interconnects. The signal posts may be connected to thermally conductive via elements and the package may include heat spreaders on upper and lower package surfaces. The first die may comprise a DRAM while the second set of die comprise portions of a general purpose microprocessor. The power signals provided to the second set of die may be connected to a capacitor terminal in the first die to provide power signal decoupling.

Robert K. Montoye

Papers

Stable SRAM cell design for the 32 nm node and beyond

An 8T-SRAM for Variability Tolerance and Low-Voltage Operation in High-Performance Caches

A 45nm CMOS neuromorphic chip with a scalable architecture for learning in networks of spiking neurons

A fully-integrated switched-capacitor 2∶1 voltage converter with regulation capability and 90% efficiency at 2.3A/mm 2

Multi-chip integrated circuit module