期刊文献+
共找到2篇文章
< 1 >
每页显示 20 50 100
Technological Exploration of RRAM Crossbar Array for Matrix-Vector Multiplication 被引量:4
1
作者 Lixue Xia Peng Gu +7 位作者 Boxun Li Tianqi Tang Xiling Yin Wenqin Huangfu shimeng yu yu Cao yu Wang Huazhong Yang 《Journal of Computer Science & Technology》 SCIE EI CSCD 2016年第1期3-19,共17页
Matrix-vector multiplication is the key operation for many computationally intensive algorithms. The emerging metal oxide resistive switching random access memory (RRAM) device and RRAM crossbar array have demonstra... Matrix-vector multiplication is the key operation for many computationally intensive algorithms. The emerging metal oxide resistive switching random access memory (RRAM) device and RRAM crossbar array have demonstrated a promising hardware realization of the analog matrix-vector multiplication with ultra-high energy efficiency. In this paper, we analyze the impact of both device level and circuit level non-ideal factors, including the nonlinear current-voltage relationship of RRAM devices, the variation of device fabrication and write operation, and the interconnect resistance as well as other crossbar array parameters. On top of that, we propose a technological exploration flow for device parameter configuration to overcome the impact of non-ideal factors and achieve a better trade-off among performance, energy, and reliability for each specific application. Our simulation results of a support vector machine (SVM) and Mixed National Institute of Standards and Technology (MNIST) pattern recognition dataset show that RRAM crossbar array based SVM is robust to input signal fluctuation but sensitive to tunneling gap deviation. A further resistance resolution test presents that a 6-bit RRAM device is able to realize a recognition accuracy around 90%, indicating the physical feasibility of RRAM crossbar array based SVM. In addition, the proposed technological exploration flow is able to achieve 10.98% improvement of recognition accuracy on the MNIST dataset and 26.4% energy savings compared with previous work. Experimental results also show that more than 84.4% power saving can be achieved at the cost of little accuracy reduction. 展开更多
关键词 resistive switching random access memory (RRAM) machine learning electronic design automation matrixvector multiplication non-ideal factor
原文传递
Profile error compensation in precision grinding of ellipsoid optical surface 被引量:3
2
作者 shimeng yu Jiahao ZHU +1 位作者 Peng YAO Chuanzhen HUANG 《Chinese Journal of Aeronautics》 SCIE EI CAS CSCD 2021年第4期115-123,共9页
Optical glass elements with the ellipsoidal surface are used in many critical aerospace instruments,such as satellites,telescope and cameras.Their optical performance is mainly affected by profile accuracy and surface... Optical glass elements with the ellipsoidal surface are used in many critical aerospace instruments,such as satellites,telescope and cameras.Their optical performance is mainly affected by profile accuracy and surface quality.In this paper,a rectangular ellipsoid surface is precisely ground on a BK7 optical glass blank by grating scanning grinding path with a three-axis CNC precision surface grinder.A profile error compensation procedure for ellipsoidal grinding is proposed based on the error analysis about the primary error sources in the XY and YZ projection planes during the grinding process.The mathematical prediction models of the wheel arc profile error,the measurement and wear error of the grinding wheel radiuses are established.By applying the proposed error compensation procedure,the profile accuracy of the ellipsoidal surface was improved from 4 lm to 2 lm in the XY plane,and improved from 15 lm to 5 lm in the YZ plane. 展开更多
关键词 Aspheric lens Ellipsoid surface Error compensation Grinding wheel truing Precision grinding
原文传递
上一页 1 下一页 到第
使用帮助 返回顶部