28 November 2017 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Stochastic computing (SC) is an unconventional method of computation that treats data as probabilities. Typically, each bit of an N-bit stochastic number (SN) X is randomly chosen to be 1 with some probability pX, and X is generated and processed by conventional logic circuits. For instance, a single AND gate performs multiplication. The value X of an SN is measured by the density of 1s in it, an information-coding scheme also found in biological neural systems. SC has uses in massively parallel systems and is very tolerant of soft errors. Its drawbacks include low accuracy, slow processing, and complex design needs. Its ability to efficiently perform tasks like communication decoding and neural network inference has rekindled interest in the field. Many challenges remain to be overcome, however, before SC becomes widespread. In this paper, we discuss the evolution of SC, mostly focusing on recent developments. We highlight the main challenges and discuss potential methods of overcoming them.
13 January 2024 in ACM Transactions on Design Automation of Electronic Systems
Given the stringent requirements of energy efficiency for Internet-of-Things edge devices, approximate multipliers, as a basic component of many processors and accelerators, have been constantly proposed and studied for decades, especially in error-resilient applications. The computation error and energy efficiency largely depend on how and where the approximation is introduced into a design. Thus, this article aims to provide a comprehensive review of the approximation techniques in multiplier designs ranging from algorithms and architectures to circuits. We have implemented representative approximate multiplier designs in each category to understand the impact of the design techniques on accuracy and efficiency. The designs can then be effectively deployed in high-level applications, such as machine learning, to gain energy efficiency at the cost of slight accuracy loss.
1 December 2021 in IEEE Transactions on Computers
Approximate computing is a promising alternative to improve energy efficiency for IoT devices on the edge. This work proposes a piecewise-linearly-approximated and unbiased floating-point approximate multiplier with run-time configurability. We provide a theoretically sound formulation that turns multiplication approximation to an optimization problem. With the formulation and findings, a multi-level architecture is proposed to easily incorporate run-time configurability and module execution parallelism. Finally, the proposed multiplier is further optimized to reduce the circuit implementation complexity, making the multiplier linearly dependent on the precision requirement, instead of quadratically or exponentially as in prior work. When compared to the prior state-of-the-art approximate floating-point multiplier, ApproxLP [1], the proposed multiplier outperforms in all the aspects including accuracy, area, and delay. By replacing a full-precision floating-point multiplier in GPU, the proposed design can improve the energy efficiency for various edge computing tasks. Even with Level 1 approximation, the proposed multiplier improves energy efficiency up to 20x for machine learning on CIFAR-10, with almost negligible accuracy loss.
25 July 2018 in IEEE Transactions on Computers
Adders are key building blocks of many error-tolerant applications. Recently, a number of approximate adders were proposed. Many of them are block-based approximate adders. For approximate circuits, besides normal metrics such as area and delay, the other important design metrics are the various error statistics, such as error rate (ER), mean error distance (MED), and mean square error (MSE). Given the popularity of block-based approximate adders, in this work, we propose an efficient method to obtain their error statistics. We first show how to calculate the ER. Then, we demonstrate an approach to get the error distribution, which can be used to calculate other metrics, such as MED and MSE. Our method is applicable to an arbitrary block-based approximate adder. It is accurate for the uniformly distributed inputs. Experimental results also demonstrated that it produces error metrics close to the accurate ones for various types of non-uniform input distributions. Compared to the state-of-the-art algorithm for obtaining the error distributions of block-based approximate adders, for the uniform input distribution, our method improves the runtime by up to $4.8\times10^{4}$ times with the same accuracy; for non-uniform input distributions, it achieves a speed-up of up to 400 times with very similar accuracy.
14 January 2021 in IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Neural network (NN) computing contains a large number of multiply-and-accumulate (MAC) operations. The performance of NN accelerator is limited with the traditional von Neumann architecture due to the tremendous off-chip memory accesses. Resistive random-access memory (ReRAM)-based crossbars can naturally perform matrix–vector multiplication (MVM) operations and are well suitable for NN accelerators. In the existing ReRAM-based NN accelerators, the synaptic weights represented by the conductances of ReRAMs are mainly based on the binary coding. However, the imperfect fabrication process combined with stochastic filament-based switching leads to resistance variations of ReRAMs, which can significantly alter the weights in binary synapses and degrade the NN accuracy. Moreover, the NN accuracy further deteriorates with multilevel cells (MLCs) used for reducing hardware overhead. In this article, a novel unary coding of synaptic weights is proposed to overcome the resistance variations of MLCs and achieve reliable ReRAM-based neuromorphic computing. A variation-aware optimal mapping scheme is also proposed in compliance with the unary coding to guarantee high accuracy by leveraging a unique feature of unary coding—the existence of multiple ways to represent the same value. The optimal mapping obtains very small errors for weights with resistance variations of MLCs. Our simulation results show that under resistance variations, the proposed method achieves less than 0.08% and 3.43% accuracy loss on CIFAR10 and ImageNet, respectively, compared to the ideal accuracy. With each synaptic weight represented by four 2-b MLCs, the proposed method improves the accuracy over the traditional binary coding scheme by 83.39% and 87.6% for CIFAR10 and ImageNet, respectively.