Low-Power Vision Transformer Accelerator With Hardware-Aware Pruning and Optimized Dataflow
Low-Power Vision Transformer Accelerator With Hardware-Aware Pruning and Optimized Dataflow
Abstract
Current transformer accelerators primarily focus on optimizing self-attention due to its quadratic complexity. However, this focus is less relevant for vision transformers with short token lengths, where the Feed-Forward Network (FFN) tends to be the dominant computational bottleneck. This paper presents a low power Vision Transformer accelerator, optimized through algorithm-hardware co-design. The model complexity is reduced using hardware-friendly dynamic token pruning without introducing complex mechanisms. Sparsity is further improved by replacing GELU with ReLU activations and employing dynamic FFN2 pruning, achieving a 61.5% reduction in operations and a 59.3% reduction in FFN2 weights, with an accuracy loss of less than 2%. The hardware adopts a rowwise dataflow with output-oriented data access to eliminate data transposition, and supports dynamic operations with minimal area overhead. Implemented in TSMC’s 28nm CMOS technology, our design occupies 496.4K gates and includes a 232KB SRAM buffer, achieving a peak throughput of 1024 GOPS at 1GHz, with an energy efficiency of 2.31 TOPS/W and an area efficiency of 858.61 GOPS/mm².
Thank you for visiting this project page. Register this project to explore its novelty and implementation support.
