Knowledge Commons of Institute of Automation,CAS
Ristretto: An Atomized Processing Architecture for Sparsity-Condensed Stream Flow in CNN | |
Gang Li1; Weixiang Xu2![]() | |
2022-07 | |
会议名称 | IEEE/ACM International Symposium on Microarchitecture |
会议日期 | 2022-10 |
会议地点 | Westin Chicago |
摘要 | Low-precision quantization and sparsity have been widely explored in CNN acceleration due to their effectiveness in reducing computational complexity and memory requirements. However, to support variable numerical precision and sparse computation, prior accelerators design flexible multipliers or sparse dataflow separately. A uniform solution that simultaneously exploits mixed-precision and dual-sided irregular sparsity for CNN acceleration is still lacking. Through an in-depth review of existing precision-scalable and sparse accelerators, we observe that a direct combination of low-level multipliers and high-level sparse dataflow from both sides is challenging due to their orthogonal design spaces. To this end, in this paper, we propose condensed streaming computation. By representing non-zero weights and activations as atomized streams, the low-level mixedprecision multiplication and high-level sparse convolution can be unified into a shared dataflow through hierarchical data reuse. Based on the condensed streaming computation, we propose Ristretto, an atomized architecture that exploits both mixedprecision and dual-sided irregular sparsity for CNN inference. We implement Ristretto in a 28nm technology node. Extensive evaluations show that Ristretto consistently outperforms three state-of-the-art CNN accelerators, including Bit Fusion, Laconic, and SparTen, in terms of performance and energy efficiency. |
七大方向——子方向分类 | AI芯片与智能计算 |
国重实验室规划方向分类 | 智能计算与学习 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/52092 |
专题 | 复杂系统认知与决策实验室_高效智能计算与学习 |
作者单位 | 1.上海交通大学 2.中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Gang Li,Weixiang Xu,Zhuoran Song,et al. Ristretto: An Atomized Processing Architecture for Sparsity-Condensed Stream Flow in CNN[C],2022. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
MICRO2022.pdf(850KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论