{"id":1082,"date":"2026-09-20T15:25:45","date_gmt":"2026-09-20T15:25:45","guid":{"rendered":"https:\/\/els-journal.net\/wp\/?page_id=1082"},"modified":"2026-09-20T15:34:47","modified_gmt":"2026-09-20T15:34:47","slug":"30203-design-of-error-efficient-inaccurate-32-compressor-for-approximate-8x8-multiplier","status":"publish","type":"page","link":"https:\/\/els-journal.net\/wp\/?page_id=1082","title":{"rendered":"30203 Design of Error Efficient Inaccurate 3:2 Compressor for Approximate 8\u00d78 Multiplier"},"content":{"rendered":"\n\n\n

Vol. 30, No. 2 – December 2026<\/h3>\n

Design of Error Efficient Inaccurate 3:2 Compressor for Approximate 8\u00d78 Multiplier<\/h3>\n

https:\/\/doi.org\/10.53314\/ELS2630074N<\/h5>\n
Srinivasa Gupta Noolu, M. Satya Anuradha, and Krishna Veni Sahukara<\/h5>\n
Abstract<\/b><\/h5>\n
Energy-efficient arithmetic units are critical for modern error-resilient computing systems. This work introduces <\/span>a novel inaccurate 3:2 compressor that exploits input probability <\/span>characteristics to reduce complexity by allowing a single, low probability <\/span>error condition. The proposed approximation alters <\/span>only the sum output for the all-one input combination, while <\/span>maintaining an exact carry output. The compressor is integrated <\/span>into a radix-4 methodology based 8\u00d78 multiplier realized through <\/span>a hierarchical composition of 2\u00d72 and 4\u00d74 multiplier modules. <\/span>Three alternative architectures are developed for intermediate <\/span>partial product accumulation using the proposed inexact 3:2 <\/span>compressor. Error behavior is analyzed in MATLAB and benchmarked <\/span>against existing inaccurate designs using Error Rate <\/span>(ER), Mean Error Distance (MED), and Normalized Mean Error <\/span>Distance (NMED). Hardware implementations are carried out in <\/span>Cadence Spectre with GPDK 45-nm technology to evaluate area, <\/span>power consumption, and delay. Experimental results indicate that <\/span>the proposed multiplier provides favorable accuracy\u2013efficiency <\/span>trade-offs. Application-level validation using image multiplication <\/span>and image convolution further demonstrates improved perceptual <\/span>quality measured through PSNR and SSIM.<\/span><\/h5>\n
Full text: <\/a><\/h5>\n\n\n\n\n\"cited<\/a>\n\n\n\n\nGoogle Scholar Citations <\/a>\n\n\n\n\n
<\/span> <\/center>