TY - GEN
T1 - An interval-based mapping algorithm for multi-shape tasks on dynamic partial reconfigurable FPGAs
AU - Zhou, Tingyu
AU - Pan, Tieyuan
AU - Meyer, Michael Conrad
AU - Dong, Yiping
AU - Watanabe, Takahiro
N1 - Funding Information:
This work was partly supported by WASEDA Tokutei-kadai 2018k-301, 2019C-719.
Publisher Copyright:
© 2020 IEEE.
PY - 2020/5
Y1 - 2020/5
N2 - The Dynamic Partial Reconfiguration (DPR) feature of reconfigurable devices allows multiple tasks to be implemented on a single device simultaneously. However, a task mapping problem emerges with the parallelism of the FPGA, especially in a real-time system, a task mapping position has to be quickly determined with a requirement of efficient usage for limited FPGA resources. Traditional task mapping algorithms simplify the task as a rectangular shape, which results in the additional internal unused areas, thus wasting a significant portion of FPGA resources. In this paper, a task mapping algorithm for the multi-shape tasks based on an interval list is proposed. Simulation results demonstrate that the FPGA utilization ratio is improved by at least 10.3% compared with existing algorithms.
AB - The Dynamic Partial Reconfiguration (DPR) feature of reconfigurable devices allows multiple tasks to be implemented on a single device simultaneously. However, a task mapping problem emerges with the parallelism of the FPGA, especially in a real-time system, a task mapping position has to be quickly determined with a requirement of efficient usage for limited FPGA resources. Traditional task mapping algorithms simplify the task as a rectangular shape, which results in the additional internal unused areas, thus wasting a significant portion of FPGA resources. In this paper, a task mapping algorithm for the multi-shape tasks based on an interval list is proposed. Simulation results demonstrate that the FPGA utilization ratio is improved by at least 10.3% compared with existing algorithms.
KW - FPGA
KW - Multi-shape tasks
KW - Reconfigurable computing
KW - Task mapping algorithm
UR - http://www.scopus.com/inward/record.url?scp=85090272233&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85090272233&partnerID=8YFLogxK
U2 - 10.1109/IPDPSW50202.2020.00028
DO - 10.1109/IPDPSW50202.2020.00028
M3 - Conference contribution
AN - SCOPUS:85090272233
T3 - Proceedings - 2020 IEEE 34th International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020
SP - 127
EP - 130
BT - Proceedings - 2020 IEEE 34th International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 34th IEEE International Parallel and Distributed Processing Symposium Workshops, IPDPSW 2020
Y2 - 18 May 2020 through 22 May 2020
ER -