Multielectrostatic FPGA Placement Considering SLICEL–SLICEM Heterogeneity, Clock Feasibility, and Timing Optimization
Jing Mai,Jiarui Wang,Zhixiong Di,Yibo Lin
DOI: https://doi.org/10.1109/tcad.2023.3313101
IF: 2.9
2023-01-01
IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems
Abstract:When modern FPGA architecture becomes increasingly complicated, modern FPGA placement is a mixed optimization problem with multiple objectives, including wirelength, routability, timing closure, and clock feasibility. Typical FPGA devices nowadays consist of heterogeneous SLICEs like SLICEL and SLICEM. The resources of a SLICE can be configured to LUT, FF, distributed RAM, SHIFT, CARRY. Besides such heterogeneity, advanced FPGA architectures also bring complicated constraints like timing, clock routing, carry chain alignment, etc. The above heterogeneity and constraints impose increasing challenges to FPGA placement algorithms. In this work, we propose a multi-electrostatic FPGA placer considering the aforementioned SLICEL-SLICEM heterogeneity under timing, clock routing and carry chain alignment constraints. We first propose an effective SLICEL-SLICEM heterogeneity model with a novel electrostatic-based density formulation. We also design a dynamically adjusted preconditioning and carry chain alignment technique to stabilize the optimization convergence. We then propose a timing-driven net weighting scheme to incorporate timing optimization. Finally, we put forward a nested Lagrangian relaxation-based placement framework to incorporate the optimization objectives of wirelength, routability, timing, and clock feasibility. Experimental results on both academic and industrial benchmarks demonstrate that our placer outperforms the state-of-the-art placers in quality and efficiency.