QForce-RL: Quantized FPGA-Optimized Reinforcement Learning Compute Engine

TL;DR

QForce-RL introduces a quantized, FPGA-optimized reinforcement learning engine that significantly improves throughput and energy efficiency while maintaining performance, suitable for resource-constrained devices.

Contribution

It presents a novel quantization-based FPGA acceleration architecture for RL, combining lightweight design with scalable deployment and performance improvements.

Findings

Performance up to 2.3x better than state-of-the-art

FPS improved by 2.6x over existing methods

Energy footprint reduced through quantization

Abstract

Reinforcement Learning (RL) has outperformed other counterparts in sequential decision-making and dynamic environment control. However, FPGA deployment is significantly resource-expensive, as associated with large number of computations in training agents with high-quality images and possess new challenges. In this work, we propose QForce-RL takes benefits of quantization to enhance throughput and reduce energy footprint with light-weight RL architecture, without significant performance degradation. QForce-RL takes advantages from E2HRL to reduce overall RL actions to learn desired policy and QuaRL for quantization based SIMD for hardware acceleration. We have also provided detailed analysis for different RL environments, with emphasis on model size, parameters, and accelerated compute ops. The architecture is scalable for resource-constrained devices and provide parametrized efficient deployment with flexibility in latency, throughput, power, and energy efficiency. The proposed QForce-RL provides performance enhancement up to 2.3x and better FPS - 2.6x compared to SoTA works.