AutoRAC: Automated Processing-in-Memory Accelerator Design for Recommender Systems

TL;DR

AutoRAC introduces an automated method for designing processing-in-memory accelerators tailored for recommender systems, significantly improving speed, area, and power efficiency over manual designs through a comprehensive search and optimization approach.

Contribution

It presents a novel automated PIM design framework that co-optimizes recommender models and hardware architecture within an extremely large design space.

Findings

Achieves up to 3.36× speedup over manual designs

Reduces area by 1.68×

Improves power efficiency by 12.48×

Abstract

The performance bottleneck of deep-learning-based recommender systems resides in their backbone Deep Neural Networks. By integrating Processing-In-Memory~(PIM) architectures, researchers can reduce data movement and enhance energy efficiency, paving the way for next-generation recommender models. Nevertheless, achieving performance and efficiency gains is challenging due to the complexity of the PIM design space and the intricate mapping of operators. In this paper, we demonstrate that automated PIM design is feasible even within the most demanding recommender model design space, spanning over $10^{54}$ possible architectures. We propose \methodname, which formulates the co-optimization of recommender models and PIM design as a combinatorial search over mixed-precision interaction operations, and parameterizes the search with a one-shot supernet encompassing all mixed-precision options. We comprehensively evaluate our approach on three Click-Through Rate benchmarks, showcasing the superiority of our automated design methodology over manual approaches. Our results indicate up to a 3.36$\times$ speedup, 1.68$\times$ area reduction, and 12.48$\times$ higher power efficiency compared to naively mapped searched designs and state-of-the-art handcrafted designs.