Browse Prior Art Database

DENSITY BASED SPARE CELL DISTRIBUTION

IP.com Disclosure Number: IPCOM000145905D
Publication Date: 2007-Jan-31
Document File: 4 page(s) / 238K

Publishing Venue

The IP.com Prior Art Database

Abstract

This paper discusses the distribution of spare cells in an integrated circuit. In our approach, the density of a region governs the number of spare modules to be placed in that region. High density regions receive more spare cells as compared to low density regions. The spare cells pertaining to a spare module are placed close together with the help of guides during placement. This density based distribution ensures uniform availability of all spare cell types in all possible areas, minimum timing impact and minimum metal layer changes required for an ECO.

This text was extracted from a Microsoft Word document.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 52% of the total text.

DENSITY BASED SPARE CELL DISTRIBUTION

 

ABSTRACT

This paper discusses the distribution of spare cells in an integrated circuit.  In our approach, the density of a region governs the number of spare modules to be placed in that region.  High density regions receive more spare cells as compared to low density regions.  The spare cells pertaining to a spare module are placed close together with the help of guides during placement.  This density based distribution ensures uniform availability of all spare cell types in all possible areas, minimum timing impact and minimum metal layer changes required for an ECO.


INTRODUCTION

An Engineering Change Order (ECO) is used to implement logical and timing design change in a taped-out database.  ECOs can arise due to design errors found during debug or changes that are made to the design specification to compensate for design problems in other areas of the system design.

ECOs are implemented by modifying only a few of the layers (typically metal layers).  This is because starting the process from the beginning will almost always require new photo masks for all layers, which is quite expensive.  A change implemented by modifying only a few metal layers is typically called a Metal ECO.

To implement Metal ECOs, there are some spare (extra) modules added to the design netlist, their number being proportional to the gate count of the design. These spare modules comprise basic gates (spare cells) that can be subsequently re-routed to correct logic or timing defects.  The method of distributing spare cells in a standard cell region of an integrated circuit plays a big role in the ECO cost.  Optimum distribution of the spare cells can reduce the number of layer changes to implement a Metal ECO.  In the design of an integrated circuit, each layer change increases non recurring engineering (NRE) cost of mask preparation.  With shrinking technologies and increasing mask costs, Metal ECOs form a substantial chunk of the design budget.

An optimal placement of spare cells in a design should ensure that once a Metal ECO is required, one should be able to find the required cell in close proximity to avoid surprises in timing and also ensure less ripping off and re-routing.

The proposed methodology offers optimal distribution of spare cells by endorsing the placement of all types of spare cells in all required regions at minimum distance from probable E...