Abstract:Multiplications have been commonly conducted in quantized CNNs, filters, and reconfigurable cores, and so on, which are widely deployed in mobile and embedded applications. Most multipliers are designed to perform multiplications with symmetric bit-widths, i.e., n - by n -bit multiplication. Such features would cause extra area overhead and performance loss when m - by n -bit multiplications ( m > n ) are deployed in the same hardware design, resulting in inefficient multiplication operations. It is highly desired and challenging to propose a reconfigurable multiplier design to accommodate operands with both symmetric and asymmetric bit-widths. In this work, we propose a reconfigurable approximate multiplier to support multiplications at various precisions, i.e., bit-widths. Unlike prior works of approximate adders assuming a uniform weight distribution with bit-wise independence, scenarios like a quantized CNN may have a centralized weight distribution and hence follow a Gaussian-like distribution with correlated adjacent bits. Thus, a new block-based approximate adder is also proposed as part of the multiplier to ensure energy-efficient operation with an awareness of the bit-wise correlation. Our experimental results show that the proposed approximate adder significantly reduces the error rate by 76% to 98% over a state-of-the-art approximate adder for Gaussian-like distribution scenarios. Evaluation results show that the proposed multiplier is 19% faster and 22% more power saving than a Xilinx multiplier IP at the same bit precision and achieves a 23.94-dB peak signal-to-noise ratio, which is comparable to the accurate one of 24.10 dB when deployed in a Gaussian filter for image processing tasks.

A New Approximate Adder with Low Relative Error and Correct Sign Calculation

Approximate Adder with Hybrid Prediction and Error Compensation Technique

A Reconfigurable Multiplier for Signed Multiplications with Asymmetric Bit-Widths.

RARA: Dataflow Based Error Compensation Methods with Runtime Accuracy-Reconfigurable Adder

A Reconfigurable Approximate Multiplier for Quantized CNN Applications.

A single clock cycle approximate adder with hybrid prediction and error compensation methods

Design and Analysis of Multiplexer based Approximate Adder for Low Power Applications

Characterizing Approximate Adders and Multipliers Optimized under Different Design Constraints

Multistage Latency Adders Architecture Employing Approximate Computing

HEAD: High-Speed Approximate HEterogeneous ADder for Error-Resilient Applications

A Review, Classification, and Comparative Evaluation of Approximate Arithmetic Circuits

Approximate Arithmetic Circuits: A Survey, Characterization, and Recent Applications

Reconfigurable lower-part approximate adder with error-tolerant application: an approach using QCA computing

A Hardware- and Accuracy-Efficient Approximate Multiplier with Error Compensation for Neural Network and Image Processing Applications

Adaptive Approximation in Arithmetic Circuits: A Low-Power Unsigned Divider Design

Energy-Efficient Approximate Full Adders Applying Memristive Serial IMPLY Logic For Image Processing

Design Methodology for Approximate Accumulator Based on Statistical Error Model.

HOAA: Hybrid Overestimating Approximate Adder for Enhanced Performance Processing Engine

Combining Structural and Timing Errors in Overclocked Inexact Speculative Adders.

A Novel Approach for Fast and Accurate Mean Error Distance Computation in Approximate Adders

Approximate Ripple Carry and Carry Lookahead Adders - A Comparative Analysis