Late Breaking Results: Leveraging Approximate Computing for Carbon-Aware DNN Accelerators

2025/02/01 by Panteleaki, Aikaterini Maria, Balaskas, Konstantinos, Zervakis, Georgios +2 · 1 citation
#FOS: Computer and information sciences #Hardware Architecture (cs.AR)

paper · doi:10.48550/arxiv.2502.00286

Abstract

The rapid growth of Machine Learning (ML) has increased demand for DNN hardware accelerators, but their embodied carbon footprint poses significant environmental challenges. This paper leverages approximate computing to design sustainable accelerators by minimizing the Carbon Delay Product (CDP). Using gate-level pruning and precision scaling, we generate area-aware approximate multipliers and optimize the accelerator design with a genetic algorithm. Results demonstrate reduced embodied carbon while meeting performance and accuracy requirements.

Cited by

• A Vertical Approach to Designing and Managing Sustainable Heterogeneous Edge Data Centers

Related

• Carbon-Efficient 3D DNN Acceleration: Optimizing Performance and Sustainability
• Energy-efficient DNN Inference on Approximate Accelerators Through Formal Property Exploration
• Design Space Exploration and Optimization for Carbon-Efficient Extended Reality Systems
• CarbonScaling: Extending Neural Scaling Laws for Carbon Footprint in Large Language Models
• CarbonCP: Carbon-Aware DNN Partitioning with Conformal Prediction for Sustainable Edge Intelligence