High-performance clock mesh optimization

Abstract

Clock meshes are extremely effective at producing low-skew regional clock networks that are tolerant of environmental and process variations. For this reason, clock meshes are used in most high-performance designs, but this robustness consumes significant power. In this work, we present two techniques to optimize high-performance clock meshes. The first technique is a mesh perturbation methodology for nonuniform mesh routing. The second technique is a skew-aware buffer placement through iterative buffer deletion. We demonstrate how these optimizations can achieve significant power reductions and a near elimination of short-circuit power. In addition, the total wire length is decreased, the number of required buffers is decreased, and both skew and robustness are improved on average when variation is considered.

Capacitance driven clock mesh synthesis to minimize skew and power dissipation

“…The general methodology which the previous works ( [2,3]) have followed is to start with a uniform mesh and increase the mesh density till the target skew is achieved. In [4], the mesh wire segments obtained by the method in [2,3] are moved horizontally or vertically such that the stub lengths are minimized. This results in a non-uniform mesh.…”

“…In cases where the distribution of clock sinks is not uniform (ispd10cns06, 07, 08), this skew-driven mesh formation is not very efficient: more wire is used, resulting in consumption of more power. Hence, in most practical cases, the mesh obtained by [2,3,4] will result in an inefficient design with more power consumption. In the proposed capacitance driven method, a non-uniform mesh is constructed based on density and capacitance of the sinks.…”

“…The mesh was not preferred since meshes were difficult to analyse and lacked automation tools. But clock mesh synthesis has received much attention since 2010 [2,3,4]. Tree driven mesh is the hybrid clock distribution scheme in which tree is used at the top level and mesh is used in the leaf level to distribute clock (Fig.…”

“…We propose 1. A "capacitance-driven" mesh formation methodology as opposed to the "skew-driven" mesh formation methodology of [2,3,4].…”

“…A new method of connecting sinks to the mesh by a combination of Steiner tree and stubs as opposed to stubs alone [2,3,4].…”