This paper describes CAVIAR, a massively parallel hardware implementation of a spike-based sensing-processing-learning-actuating system inspired by the physiology of the nervous system. CAVIAR uses the asynchronous address-event representation (AER) communication framework and was developed in the context of a European Union funded project. It has four custom mixed-signal AER chips, five custom digital AER interface components, 45 k neurons (spiking cells), up to 5 M synapses, performs 12 G synaptic operations per second, and achieves millisecond object recognition and tracking latencies.

Event-Driven vision sensing is a new way of sensing visual reality in a frame-free manner. This is, the vision sensor (camera) is not capturing a sequence of still frames, as in conventional video and computer vision systems. In Event-Driven sensors each pixel autonomously and asynchronously decides when to send its address out. This way, the sensor output is a continuous stream of address events representing reality dynamically continuously and without constraining to frames. In this paper we present an Event-Driven Convolution Module for computing 2D convolutions on such event streams. The Convolution Module has been designed to assemble many of them for building modular and hierarchical Convolutional Neural Networks for robust shape and pose invariant object recognition. The Convolution Module has multi-kernel capability. This is, it will select the convolution kernel depending on the origin of the event. A proof-of-concept test prototype has been fabricated in a 0.35 μm CMOS process and extensive experimental results are provided. The Convolution Processor has also been combined with an Event-Driven Dynamic Vision Sensor (DVS) for high-speed recognition examples. The chip can discriminate propellers rotating at 2 k revolutions per second, detect symbols on a 52 card deck when browsing all cards in 410 ms, or detect and follow the center of a phosphor oscilloscope trace rotating at 5 KHz.

We present a neuromorphic cortical-layer processing microchip for address event representation (AER) spike-based processing systems. The microchip computes 2-D convolutions of video information represented in AER format in real time. AER, as opposed to conventional frame-based video representation, describes visual information as a sequence of events or spikes in a way similar to biological brains. This format allows for fast information identification and processing, without waiting to process complete image frames. The neuromorphic cortical-layer processing microchip presented in this paper computes convolutions of programmable kernels over the AER visual input information flow. It not only computes convolutions but also allows for a programmable forgetting rate, which in turn allows for a bio-inspired coincidence detection processing. Kernels are programmable and can be of arbitrary shape and arbitrary size of up to 32 times 32 pixels. The convolution processor operates on a pixel array of size 32 times 32, but can process an input space of up to 128 times 128 pixels. Larger pixel arrays can be directly processed by tiling arrays of chips. The chip receives and generates data in AER format, which is asynchronous and digital. However, its internal operation is based on analog low-current circuit techniques. The paper describes the architecture of the chip and circuits used for the pixels, including calibration techniques to overcome mismatch. Extensive experimental results are provided, describing pixel operation and calibration, convolution processing with and without forgetting, and high-speed recognition experiments like discriminating rotating propellers of different shape rotating at speeds of up to 5000 revolutions per second

In this paper, a chip that performs real-time image convolutions with programmable kernels of arbitrary shape is presented. The chip is a first experimental prototype of reduced size to validate the implemented circuits and system level techniques. The convolution processing is based on the address-event-representation (AER) technique, which is a spike-based biologically inspired image and video representation technique that favors communication bandwidth for pixels with more information. As a first test prototype, a pixel array of 16times16 has been implemented with programmable kernel size of up to 16times16. The chip has been fabricated in a standard 0.35 mum complimentary metal-oxide-semiconductor (CMOS) process. The technique also allows to process larger size images by assembling 2D arrays of such chips. Pixel operation exploits low-power mixed analog-digital circuit techniques. Because of the low currents involved (down to nanoamperes or even picoamperes), an important amount of pixel area is devoted to mismatch calibration. The rest of the chip uses digital circuit techniques, both synchronous and asynchronous. The fabricated chip has been thoroughly tested, both at the pixel level and at the system level. Specific computer interfaces have been developed for generating AER streams from conventional computers and feeding them as inputs to the convolution chip, and for grabbing AER streams coming out of the convolution chip and storing and analyzing them on computers. Extensive experimental results are provided. At the end of this paper, we provide discussions and results on scaling up the approach for larger pixel arrays and multilayer cortical AER systems.

Cryptocircuits can be attacked by third parties using differential power analysis (DPA), which uses power consumption dependence on data being processed to reveal critical information. To protect security devices against this issue, differential logic styles with (almost) constant power dissipation are widely used. However, to use such circuits effectively for secure applications it is necessary to eliminate any energy-secure flaw in security in the shape of memory effects that could leak information. This paper proposes a design methodology to improve pull-down logic configuration for secure differential gates by redistributing the charge stored in internal nodes and thus, removing memory effects that represent a significant threat to security. To evaluate the methodology, it was applied to the design of AND/NAND and XOR/XNOR gates in a 90 nm technology, adopting the sense amplifier based logic (SABL) style for the pull-up network. The proposed solutions leak less information than typical SABL gates, increasing security by at least two orders of magnitude and with negligible performance degradation. A simulation-based DPA attack on the Sbox9 cryptographic module used in the Kasumi algorithm, implemented with complementary metal-oxide-semiconductor, SABL and proposed gates, was performed. The results obtained illustrate that the number of measurements needed to disclose the key increased by much more than one order of magnitude when using our proposal. This paper also discusses how the effectivenness of DPA attacks is influenced by operating temperature and details how to insure energy-secure operations in the new proposals.

Differential Fault Analysis (DFA) and Power Analysis (PA) attacks, have become the main methods for exploiting the vulnerabilities of physical implementations of block ciphers, currently used in a multitude of applications, such as the Advanced Encryption Standard (AES). In order to minimize these types of vulnerabilities, several mechanisms have been proposed to detect fault attacks. However, these mechanisms can have a significant cost, not fully covering the implementations against fault attacks or not taking into account the leakage of the information exploitable by the power analysis attacks. In this paper, four different approaches are proposed with the aim of protecting the AES block cipher against DFA. The proposed solutions are based on Hamming code and parity bits as signature generators for the internal state of the AES cipher. These allow to detect DFA exploitable faults, from bit to byte level. The proposed solutions have been applied to a T-box based AES block cipher implemented on Field Programmable Gate Array (FPGA). Experimental results suggest a fault coverage of 98.5% and 99.99% with an area penalty of 9% and 36% respectively, for the parity bit signature generators and a fault coverage of 100% with an area penalty of 18% and 42% respectively when Hamming code signature generator is used. In addition, none of the proposed countermeasures impose a frequency degradation, in respect to the unprotected cipher. The proposed work goes further in the evaluation of the proposed DFA countermeasures by evaluating the impact of these structures in terms of power side-channel. The obtained results suggest that no extra information leakage is produced that can be exploited by PA. Overall, the proposed DFA countermeasures provide a high fault coverage protection with a low cost in terms of area and power consumption and no PA security degradation.