HFRWKV: A High-Performance Fully On-Chip Hardware Accelerator for RWKV

TL;DR

HFRWKV is a specialized FPGA-based hardware accelerator for RWKV, a modern RNN that balances long-context processing with efficient hardware utilization, achieving significant throughput and energy efficiency improvements.

Contribution

The paper introduces HFRWKV, a fully on-chip FPGA accelerator with novel hybrid-precision quantization and optimized complex operation modules for RWKV.

Findings

63.48× throughput improvement over CPU

139.17× energy efficiency over CPU

32.33× throughput improvement over GPU

Abstract

RWKV is a modern RNN architecture that approaches the performance of Transformers, with the advantage of processing long contexts at a linear memory cost. However, its sequential computation pattern struggles to efficiently leverage GPU parallelism, which leads to low compute resource utilization. Furthermore, frequent off-chip weight accesses create a memory bottleneck. To address these challenges, we propose HFRWKV, an FPGA-based hardware accelerator specifically designed for RWKV. Within the matrix operation module, we propose a novel hardware-friendly hybrid-precision quantization strategy, which enhances performance while maintaining acceptable accuracy. For the complex operations including exponentiation and division, we introduce a method featuring reusable architectures combined with lookup tables or piecewise linear approximation, which is algorithmically refined to effectively balance precision and hardware resource consumption. Based on this foundation, we adopt a fully on-chip computing system integrating parallel matrix-vector processing array and an efficient pipeline architecture. Through computation reordering and chunked double buffering, it effectively eliminates data transfer bottlenecks and improves overall throughput. We implement HFRWKV on the Alveo U50 and U280 platform. Experimental results show that compared to a CPU, a throughput improvement of 63.48$\times$ and an energy efficiency improvement of 139.17$\times$. Compared to GPUs, achieves a throughput improvement of 32.33$\times$ and an energy efficiency improvement of 171.36$\times$.