Search papers, labs, and topics across Lattice.
This paper introduces a GPU-accelerated framework for analyzing effective resistance in 3D integrated circuit power delivery networks (PDNs), addressing the challenges posed by increased current demands and design complexity. By optimizing the planning of through-silicon vias (TSVs), the framework significantly enhances the reliability of PDNs, which is crucial for preventing issues like uneven effective resistance and IR drop. The proposed method achieves a remarkable speedup of 5 to 6 orders of magnitude over traditional solvers while ensuring minimal error in resistance calculations.
Achieving a 5 to 6 orders of magnitude speedup in effective resistance analysis could revolutionize the design of reliable 3D IC power delivery networks.
Three-dimensional (3D) integration is a critical technique for enhancing transistor density, improving power efficiency, and reducing interconnect delays. However, as current demands and design complexity increase, power deliver networks (PDNs) are facing growing challenges.Careful planning of through-silicon vias (TSVs) is essential for ensuring reliable PDNs, where effective resistance serves as a vital metric for the reliability. Ill-planned TSVs often cause 3D IC with unevenly distributed effective resistance and consequently severer IR Drop.In this paper, we propose a GPU-accelerated framework on accurate effective resistance analysis for early stage 3D IC PDNs. The proposed framework achieves a speedup of 5 to 6 orders of magnitude compared to the conventional direct solver, while maintaining negligible deviations in both maximum and average relative errors.
