A Unified Framework for Mapping and Synthesis of Approximate R-Blocks CGRAs

TL;DR

This paper presents a comprehensive framework for integrating approximate multiplication components into CGRA architectures, optimizing energy efficiency while maintaining accuracy for neural network inference.

Contribution

It introduces a novel exploration and synthesis methodology for approximate CGRAs, enabling energy-efficient design with minimal accuracy loss.

Findings

Achieves 30% power reduction with 2% area increase.

Delivers up to 440 GOPS/W on MobileNetV2 with small error.

Outperforms state-of-the-art CGRA architectures in throughput and energy efficiency.

Abstract

The ever-increasing complexity and operational diversity of modern Neural Networks (NNs) have caused the need for low-power and, at the same time, high-performance edge devices for AI applications. Coarse Grained Reconfigurable Architectures (CGRAs) form a promising design paradigm to address these challenges, delivering a close-to-ASIC performance while allowing for hardware programmability. In this paper, we introduce a novel end-to-end exploration and synthesis framework for approximate CGRA processors that enables transparent and optimized integration and mapping of state-of-the-art approximate multiplication components into CGRAs. Our methodology introduces a per-channel exploration strategy that maps specific output features onto approximate components based on accuracy degradation constraints. This enables the optimization of the system's energy consumption while retaining the accuracy above a certain threshold. At the circuit level, the integration of approximate components enables the creation of voltage islands that operate at reduced voltage levels, which is attributed to their inherently shorter critical paths. This key enabler allows us to effectively reduce the overall power consumption by an average of 30% across our analyzed architectures, compared to their baseline counterparts, while incurring only a minimal 2% area overhead. The proposed methodology was evaluated on a widely used NN model, MobileNetV2, on the ImageNet dataset, demonstrating that the generated architectures can deliver up to 440 GOPS/W with relatively small output error during inference, outperforming several State-of-the-Art CGRA architectures in terms of throughput and energy efficiency.