Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper introduces a novel placement flow specifically designed for 3D FPGAs, addressing the limitations of existing 2D-centric tools in capturing the unique characteristics of 3D fabrics. The flow incorporates partitioning-based initialization, adaptive cost scheduling, refined delay estimation, and a simulated annealing move set tailored for 3D architectures. Experimental results demonstrate significant reductions in critical-path delay and routed wirelength compared to the Verilog-To-Routing (VTR) baseline across four different 3D FPGA architectures.
3D FPGAs can finally realize their performance potential, thanks to a new placement flow that understands their unique vertical integration.
3D field-programmable gate arrays (FPGAs) promise higher performance through vertical integration. However, existing placement tools, largely inherited from 2D frameworks, fail to capture the unique delay characteristics and optimization dynamics of 3D fabrics. We introduce a 3D FPGA placement flow that integrates partitioning-based initialization, adaptive cost scheduling, refined delay estimation, and a simulated annealing move set -- all targeted at 3D FPGA architecture. Together, these enhancements improve timing estimates and the exploration of layer assignments during placement. Compared to Verilog-To-Routing (VTR), our experiments show geometric-mean (max) critical-path delay reductions of ~3% (~7%), ~2% (~4%), ~3% (~8%), and ~6% (~18%) for four 3D architectures: 3D CB, 3D CB-O, 3D CB-I, and 3D SB, respectively. We also achieve geometric-mean (max) routed wirelength reductions of ~1% (~3%), ~2% (~8%),<1% (~5%), and ~5% (~10%), respectively. Our work will be permissively open-sourced on GitHub.