Abstract
Recent studies have shown, contrary to what was previously believed, that by exploiting correlation in stochastic computing (SC) designs, more accurate SC circuits with low area cost can be realized. However, if these basic SC circuits or blocks are cascaded in series to form a large complex system, correlation between stochastic numbers (SNs) from one block to the next would be lost; thus, inaccuracies are introduced. In this study, we propose correlating circuits to be used in building complex correlated SC systems. One of the circuits is the correlator that restores lost correlations between two SNs due to previous processing. In addition, a correlated SN generator is introduced to generate SN correlated to a specific SN. Experimental results show that our methods have improved the accuracy of stochastic computation and preserved the stochastic computing correlation without the need for conversion from SC to the conventional binary-encoded computing, and vice versa. Consequently, lower latency and lower area cost are achieved.
Highlights
Stochastic computing (SC) is a reemerging computing paradigm that was first introduced by Gaines [1] in the 1960s as an alternative to the conventional binary-encoded deterministic computing (DC) technique
This paper proposes a methodology to build a design in which correlation across a correlated stochastic computing (CSC) system is maintained without utilizing multiple conversion circuits
Our work extends the categorization further to show that some elements have an effect on correlation and certain methods are proposed to allow cascading circuits to utilize correlation or create complex CSC circuits without accuracy loss, the correlation will be lost
Summary
Stochastic computing (SC) is a reemerging computing paradigm that was first introduced by Gaines [1] in the 1960s as an alternative to the conventional binary-encoded deterministic computing (DC) technique. An image processing pipeline would consist of a series of circuits that include a median filter block, a smoothing filter, an edge detector, and, a thresholding stage Such a system could not be realized previously, because the SC-based functional blocks with correlated input bit-stream produced uncorrelated output bit-streams. The designer inserts conversion circuits whenever inputs have to be correlated to restore any lost correlation This solution is infeasible since it introduces long conversion latency and significantly increases area cost (or resource utilization). Our contributions are as follows: 1) An SN correlator that restores the lost correlation between SNs and eliminates the need for conversion circuits between the functional blocks
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: TURKISH JOURNAL OF ELECTRICAL ENGINEERING & COMPUTER SCIENCES
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
