Free space management for cut-based placement

  • Authors:
  • Charles J. Alpert;Gi-Joon Nam;Paul G. Villarrubia

  • Affiliations:
  • IBM Corporation, Austin, TX;IBM Corporation, Austin, TX;IBM Corporation, Austin, TX

  • Venue:
  • Proceedings of the 2002 IEEE/ACM international conference on Computer-aided design
  • Year:
  • 2002

Quantified Score

Hi-index 0.01

Visualization

Abstract

IP blocks and large macro cells are increasingly prevalent in physical design, actually causing an increase in the available free space for the dust logic. We observe that top-down placement based on recursive bisection with multilevel partitioning performs poorly on these porous designs. However, analytic solvers have the ability to find the natural distribution of cells in the layout. Consequently, we propose an enhancement to cut-based placement called Analytic Constraint Generation (ACG). ACG utilizes an analytic engine to set constraints for the multi-level partitioner. We show that for real industry designs, ACG significantly improves the performance of cut-based placement, as implemented within a state-of-the-art industrial placer.