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.

AxBMs: Approximate Radix-8 Booth Multipliers for High-Performance FPGA-Based Accelerators

AxBMs: Approximate Radix-8 Booth Multipliers for High-Performance FPGA-Based Accelerators

Approximate Radix-8 Booth Multipliers for Low-Power and High-Performance Operation

Approximate radix-8 Booth multiplier for low power and high speed applications

Design of Approximate Radix-4 Booth Multipliers for Error-Tolerant Computing

Optimally Approximated and Unbiased Floating-Point Multiplier with Runtime Configurability

High Speed Low Power Radix 4 Approximate Booth Multiplier

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

Energy-Efficient Approximate Floating-Point Multiplier Based on Radix-8 Booth Encoding

Design of floating point multiplier using approximate hybrid Radix-4/ Radix-8 booth encoder for image analysis

Design and Implementation of Area Efficient Low Latency Radix-8 Multiplier on FPGA

Implementation of Booth Multiplier Algorithm using Radix-4 in FPGA

FPGA Implementation of an Approximate Booth-Wallace Tree Multiplier

Asic Implementation of 12-Bit Radix-8 Booth Multiplier

Implementation and Comparison of Radix-8 Booth Multiplier by using 32-bit Parallel prefix adders for High Speed Arithmetic Applications

A Hybrid Multiplier Architecture Using Partially Redundant Booth Algorithm

Efficient implementation of signed multipliers on FPGAs

Design and Implementation of Radix 8 Booth Encoding Multiplier for Low Area and High-Speed Applications

FHAM: FPGA-based High-Efficiency Approximate Multipliers Via LUT Encoding

Design of Majority Logic-Based Approximate Booth Multipliers for Error-Tolerant Applications

Design and implementation of hybrid (radix-8 Booth and TRAM) approximate multiplier using 15-4 approximate compressors for image processing application