LightMamba: Efficient Mamba Acceleration on FPGA with Quantization and Hardware Co-design
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866909834443489280 |
|---|---|
| author | Wei, Renjie Xu, Songqiang Zhong, Linfeng Yang, Zebin Guo, Qingyu Wang, Yuan Wang, Runsheng Li, Meng |
| author_facet | Wei, Renjie Xu, Songqiang Zhong, Linfeng Yang, Zebin Guo, Qingyu Wang, Yuan Wang, Runsheng Li, Meng |
| contents | State space models (SSMs) like Mamba have recently attracted much attention. Compared to Transformer-based large language models (LLMs), Mamba achieves linear computation complexity with the sequence length and demonstrates superior performance. However, Mamba is hard to accelerate due to the scattered activation outliers and the complex computation dependency, rendering existing LLM accelerators inefficient. In this paper, we propose LightMamba that co-designs the quantization algorithm and FPGA accelerator architecture for efficient Mamba inference. We first propose an FPGA-friendly post-training quantization algorithm that features rotation-assisted quantization and power-of-two SSM quantization to reduce the majority of computation to 4-bit. We further design an FPGA accelerator that partially unrolls the Mamba computation to balance the efficiency and hardware costs. Through computation reordering as well as fine-grained tiling and fusion, the hardware utilization and memory efficiency of the accelerator get drastically improved. We implement LightMamba on Xilinx Versal VCK190 FPGA and achieve 4.65x to 6.06x higher energy efficiency over the GPU baseline. When evaluated on Alveo U280 FPGA, LightMamba reaches 93 tokens/s, which is 1.43x that of the GPU baseline. Our code is available at https://github.com/PKU-SEC-Lab/LightMamba. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2502_15260 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LightMamba: Efficient Mamba Acceleration on FPGA with Quantization and Hardware Co-design Wei, Renjie Xu, Songqiang Zhong, Linfeng Yang, Zebin Guo, Qingyu Wang, Yuan Wang, Runsheng Li, Meng Computation and Language State space models (SSMs) like Mamba have recently attracted much attention. Compared to Transformer-based large language models (LLMs), Mamba achieves linear computation complexity with the sequence length and demonstrates superior performance. However, Mamba is hard to accelerate due to the scattered activation outliers and the complex computation dependency, rendering existing LLM accelerators inefficient. In this paper, we propose LightMamba that co-designs the quantization algorithm and FPGA accelerator architecture for efficient Mamba inference. We first propose an FPGA-friendly post-training quantization algorithm that features rotation-assisted quantization and power-of-two SSM quantization to reduce the majority of computation to 4-bit. We further design an FPGA accelerator that partially unrolls the Mamba computation to balance the efficiency and hardware costs. Through computation reordering as well as fine-grained tiling and fusion, the hardware utilization and memory efficiency of the accelerator get drastically improved. We implement LightMamba on Xilinx Versal VCK190 FPGA and achieve 4.65x to 6.06x higher energy efficiency over the GPU baseline. When evaluated on Alveo U280 FPGA, LightMamba reaches 93 tokens/s, which is 1.43x that of the GPU baseline. Our code is available at https://github.com/PKU-SEC-Lab/LightMamba. |
| title | LightMamba: Efficient Mamba Acceleration on FPGA with Quantization and Hardware Co-design |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2502.15260 |