Affiliation:
1. Politecnico di Milano, Italy
2. Pacific Northwest National Laboratory, USA
Abstract
This paper presents a methodology for the Synthesis of PARallel multi-Threaded Accelerators (SPARTA) from OpenMP annotated C/C++ specifications. SPARTA extends an open-source HLS tool, enabling the generation of accelerators that provide latency tolerance for irregular memory accesses through multithreading, support fine-grained memory-level parallelism through a hot-potato deflection-based network-on-chip (NoC), support synchronization constructs, and can instantiate memory-side caches. Our approach is based on a custom runtime OpenMP library, providing flexibility and extensibility. Experimental results show high scalability when synthesizing irregular graph kernels. The accelerators generated with our approach are, on average, 2.29x faster than state-of-the-art HLS methodologies.
Publisher
Association for Computing Machinery (ACM)
Reference62 articles.
1. 2023. Graph500. https://graph500.org
2. HLS-based High-throughput and Work-efficient Synthesizable Graph Processing Template Pipeline
3. A scalable processing-in-memory accelerator for parallel graph processing
4. ARM Developers. 2020. AMBA AXI and ACE Protocol Specification. https://developer.arm.com/documentation/ihi0022/e/AMBA-AXI3-and-AXI4-Protocol-Specification
5. Abanti Basak, Shuangchen Li, Xing Hu, Sang Min Oh, Xinfeng Xie, Li Zhao, Xiaowei Jiang, and Yuan Xie. 2019. Analysis and Optimization of the Memory Hierarchy for Graph Processing Workloads. In IEEE International Symposium on High Performance Computer Architecture (HPCA). 373–386.