TY - GEN
T1 - Fast and effective placement and routing directed high-level synthesis for FPGAs
AU - Zheng, Hongbin
AU - Gurumani, Swathi T.
AU - Rupnow, Kyle
AU - Chen, Deming
N1 - Copyright:
Copyright 2014 Elsevier B.V., All rights reserved.
PY - 2014
Y1 - 2014
N2 - Achievable frequency (fmax) is a widely used input con-straint for designs targeting Field-Programmable Gate Ar-rays (FPGA), because of its impact on design latency and throughput. fmax is limited by critical path delay, which is highly in uenced by lower-level details of the circuit imple-mentation such as technology mapping, placement and rout-ing. However, for high-level synthesis (HLS) design ows, it is challenging to evaluate the real critical delay at the behavioral level. Current HLS flows typically use module pre-characterization for delay estimates. However, we will demonstrate that such delay estimates are not sufficient to obtain high fmax and also minimize total execution latency. In this paper, we introduce a new HLS flow that inte-grates with Altera's Quartus synthesis and fast placement and routing (PAR) tool to obtain realistic post-PAR delay estimates. This integration enables an iterative flow that im-proves the performance of the design with both behavioral-level and circuit-level optimizations using realistic delay in-formation. We demonstrate our HLS flow produces up to 24% (on average 20%) improvement in fmax and upto 22% (on average 20%) improvement in execution latency. Fur-thermore, results demonstrate that our flow is able to achieve from 65% to 91% of the theoretical fmax on Stratix IV de-vices (550MHz).
AB - Achievable frequency (fmax) is a widely used input con-straint for designs targeting Field-Programmable Gate Ar-rays (FPGA), because of its impact on design latency and throughput. fmax is limited by critical path delay, which is highly in uenced by lower-level details of the circuit imple-mentation such as technology mapping, placement and rout-ing. However, for high-level synthesis (HLS) design ows, it is challenging to evaluate the real critical delay at the behavioral level. Current HLS flows typically use module pre-characterization for delay estimates. However, we will demonstrate that such delay estimates are not sufficient to obtain high fmax and also minimize total execution latency. In this paper, we introduce a new HLS flow that inte-grates with Altera's Quartus synthesis and fast placement and routing (PAR) tool to obtain realistic post-PAR delay estimates. This integration enables an iterative flow that im-proves the performance of the design with both behavioral-level and circuit-level optimizations using realistic delay in-formation. We demonstrate our HLS flow produces up to 24% (on average 20%) improvement in fmax and upto 22% (on average 20%) improvement in execution latency. Fur-thermore, results demonstrate that our flow is able to achieve from 65% to 91% of the theoretical fmax on Stratix IV de-vices (550MHz).
KW - High-Level Synthesis
KW - Layout driven
KW - Scheduling
UR - http://www.scopus.com/inward/record.url?scp=84899012934&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84899012934&partnerID=8YFLogxK
U2 - 10.1145/2554688.2554775
DO - 10.1145/2554688.2554775
M3 - Conference contribution
AN - SCOPUS:84899012934
SN - 9781450326711
T3 - ACM/SIGDA International Symposium on Field Programmable Gate Arrays - FPGA
SP - 1
EP - 10
BT - FPGA 2014 - Proceedings of the 2014 ACM/SIGDA International Symposium on Field Programmable Gate Arrays
PB - Association for Computing Machinery
T2 - 2014 ACM/SIGDA International Symposium on Field Programmable Gate Arrays, FPGA 2014
Y2 - 26 February 2014 through 28 February 2014
ER -