Indirect connection aware attraction for FPGA clustering (abstract only)

Authors:
Meng Yang;Jiarong Tong;A.E.A. Almaini
Affiliations:
State Key Lab of ASIC, Fudan University, Shanghai, China;State Key Lab of ASIC, Fudan University, Shanghai, China;Edinburgh Napier University, Edinburgh, United Kingdom
Venue:
Proceedings of the ACM/SIGDA international symposium on Field programmable gate arrays
Year:
2013

Citing 6
Cited 0

Using cluster-based logic blocks and timing-driven packing to improve FPGA speed and density

FPGA '99 Proceedings of the 1999 ACM/SIGDA seventh international symposium on Field programmable gate arrays
Efficient circuit clustering for area and power reduction in FPGAs

ACM Transactions on Design Automation of Electronic Systems (TODAES)
VPR: A new packing, placement and routing tool for FPGA research

FPL '97 Proceedings of the 7th International Workshop on Field-Programmable Logic and Applications
VPR 5.0: FPGA cad and architecture exploration tools with single-driver routing, heterogeneity and process scaling

Proceedings of the ACM/SIGDA international symposium on Field programmable gate arrays
MO-pack: many-objective clustering for FPGA CAD

Proceedings of the 48th Design Automation Conference
The VTR project: architecture and CAD for FPGAs from verilog to routing

Proceedings of the ACM/SIGDA international symposium on Field Programmable Gate Arrays

Quantified Score

Hi-index	0.00

Visualization

Abstract

Indirect connection aware attraction clustering algorithm is proposed for clustered field programmable gate array architecture model to achieve simultaneously optimization of several performance metrics. A new cost function considers the attraction of the subsequent basic logic elements (BLEs) to the selected cluster, the number of the used pins already in the cluster, as well as critical path delay. The attractions of which BLEs are directly and indirectly connected to the selected cluster are taken into account. As a result, more external nets are absorbed into clusters, less number of pins per cluster and fewer clusters are required. Hence, smaller channel width is required for routing and speed of the design is improved. Performance comparisons are carried out in details with respect to state-of-the-art clustering techniques interconnect resource aware clustering (iRAC) and many-objective clustering (MO-Pack). Results show that the proposed algorithm outperforms these two clustering approaches with achievements of 38.8% and 42.2% respectively in terms of channel widths and 40.1% and 44.8% respectively in terms of number of external nets but with no critical path and area overhead.