VecPAC: A Vectorizable and Precision-Aware CGRA

Conference · Thu Nov 30 04:00:00 EST 2023

DOI:https://doi.org/10.1109/ICCAD57390.2023.10323910· OSTI ID:2348882

Tan, Cheng ^[1]; Patil, Deepak ^[2]; Tumeo, Antonino ^[3]; Weisz, Gabriel ^[4]; Reinhardt, Steve ^[4]; Zhang, Jeff ^[5]

---

1. Google
2. Arizona State University
3. BATTELLE (PACIFIC NW LAB)
4. Microsoft
5. Harvard University

This paper proposes VecPAC -- a vectorizable and precision-aware coarse-grained reconfigurable array (CGRA) design. VecPAC integrates CGRA tiles with scalar functional units and specialized tiles with vector functional units that can trade off the number of vector lanes for the accuracy of the computation. We discuss the architecture design and present the related compilation framework. The experimental evaluation on a set of applications from three different domains (embedded, machine learning, and high-performance computing) shows that the hybrid design of VecPAC outperforms CGRAs with only scalar functional units by 1.48x, while providing higher scalability.

• Research Organization: Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC05-76RL01830
• OSTI ID: 2348882
• Report Number(s): PNNL-SA-180016
• Country of Publication: United States
• Language: English