WindMill: A Parameterized and Pluggable CGRA Implemented by DIAG Design Flow

TL;DR

This paper introduces WindMill, a flexible, parameterized CGRA generator built with the DIAG design flow, enabling rapid customization and achieving significant performance gains over GPUs in reinforcement learning tasks.

Contribution

The paper presents WindMill, a novel plug-in based CGRA generator, and the DIAG design flow, enhancing flexibility, reducing design complexity, and accelerating hardware development.

Findings

WindMill enables agile hardware accelerator generation.

Reinforcement learning tasks see 2.3x performance over GPU.

The DIAG flow improves design flexibility and productivity.

Abstract

With the cross-fertilization of applications and the ever-increasing scale of models, the efficiency and productivity of hardware computing architectures have become inadequate. This inadequacy further exacerbates issues in design flexibility, design complexity, development cycle, and development costs (4-d problems) in divergent scenarios. To address these challenges, this paper proposed a flexible design flow called DIAG based on plugin techniques. The proposed flow guides hardware development through four layers: definition(D), implementation(I), application(A), and generation(G). Furthermore, a versatile CGRA generator called WindMill is implemented, allowing for agile generation of customized hardware accelerators based on specific application demands. Applications and algorithm tasks from three aspects is experimented. In the case of reinforcement learning algorithm, a significant performance improvement of $2.3\times$ compared to GPU is achieved.