The quantization of Large Language Models (LLMs) poses significant challenges due to the heterogeneous nature of feature point distributions in low-bit quantization scenarios, including salient points, normal outliers, and massive outliers. These challenges are particularly pronounced in supporting both weight-only and weight-activation quantization modes, as existing methods often focus on a single mode and fail to address the diverse feature characteristics holistically, resulting in suboptimal model accuracy and hardware efficiency trade-offs.To tackle these limitations, we introduce Amove, a novel co-design framework that synergistically integrates data type and hardware architecture design for efficient LLM quantization. Our approach is threefold: First, we conduct a comprehensive analysis of quantization granularity and propose a residual approximation mechanism that balances model accuracy and memory overhead under fine-grained quantization. Second, we design a flexible fine-grained grouped vectorized data type, enabling seamless support for both weight-activation and low-bit weight-only quantization modes within a unified framework. Third, we implement the hardware architecture of Amove on both GPU tensor core and systolic array-based architectures. The Amove-enhanced tensor core achieves an average speedup of 2.13× and a 1.70× reduction in energy consumption over the state-of-the-art OliVe design. Furthermore, an Amove-based accelerator achieves up to 2.67× speedup and 1.68× energy reduction over the state-of-the-art accelerator.
Xie et al. (Fri,) studied this question.