Teo Ene

09/23/2020, 2:10 PM
li1 <-> met1 rail contacts: We based our decision here based on prior experience working with cells that have the lowest 2 metal layers running in parallel directions (such as 12LP). Since the power rails abut rather than overlap, we wanted to offer the PnR tools that use our cells the chance to place vias in the middle of the 0.61um power rails that results from the abutment of the cells' 0.305um half-rails. I had also had a bad experience involving m1 <-> m2 rail vias on 12LP, so I just sought to offer maximum flexibility to the PnR toolflow. Throughout our testing we have instructed the PnR tool to add vias in the middle of the rails. This decision was made at the very start of our design process, before we became fully aware of most of the other interconnected parts of the 130nm Open PDK project.