Optimal aspect ratios of building blocks in VLSI
Shmuel Wimer, Israel Koren, I. Cederbaum
Abstract
Shmuel Wimer, Israel Koren, I. Cederbaum
Abstract
The building blocks in a given floor-plan may have several possible physical implementations yie1din.g dif-ferent layouts. This paper discusses the problem of selecting an optimal implementation for each building block so that the area of the final layout is minimized. A polynomial algorithm that solves this problem for slicing floorplans was presented elsewhere, and it has been proved that for general (non-slicing) floorplans the problem is NP-complete. We suggest a bfranch and bound algorithm which proves to be very eflicient and can handle successfully large general non-slicing floorplans. We show also how the non-slicing and the slicing algorithms can be combined to handle effi-ciently very large general floorplans. 1.
OpenAlex reports 23 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
The building blocks in a given floor-plan may have several possible physical implementations yie1din.g dif-ferent layouts. This paper discusses the problem of selecting an optimal implementation for each building block so that the area of the final layout is minimized. A polynomial algorithm that solves this problem for slicing floorplans was presented elsewhere, and it has been proved that for general (non-slicing) floorplans the problem is NP-complete. We suggest a bfranch and bound algorithm which proves to be very eflicient and can handle successfully large general non-slicing floorplans. We show also how the non-slicing and the slicing algorithms can be combined to handle effi-ciently very large general floorplans. 1.
Key concepts: Slicing, Floorplan, Computer science, Very-large-scale integration, Block (permutation group theory), Upper and lower bounds, Integrated circuit layout, Algorithm