Crosstalk Noise Verification in Digital Designs with Interconnect Process Variations
VLSID '01 Proceedings of the The 14th International Conference on VLSI Design (VLSID '01)
Impact of Process Variations on Multi-Level Signaling for On-Chip Interconnects
VLSID '05 Proceedings of the 18th International Conference on VLSI Design held jointly with 4th International Conference on Embedded Systems Design
Stochastic analysis of interconnect performance in the presence of process variations
Proceedings of the 2004 IEEE/ACM International conference on Computer-aided design
ICCAD '05 Proceedings of the 2005 IEEE/ACM International conference on Computer-aided design
Pessimism reduction in crosstalk noise aware STA
ICCAD '05 Proceedings of the 2005 IEEE/ACM International conference on Computer-aided design
Parallel vs. serial on-chip communication
Proceedings of the 2008 international workshop on System level interconnect prediction
Two-phase synchronization with sub-cycle latency
Integration, the VLSI Journal
Is overlay error more important than interconnect variations in double patterning?
Proceedings of the 11th international workshop on System level interconnect prediction
Asynchronous current mode serial communication
IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Hi-index | 0.00 |
Manufacturable design requires matching of interconnects which have equal nominal dimensions. New design rules are projected to bring guarantee rules for interconnect matching. In this paper, we present a methodology to generate additional guarantees given the limited set of guarantees in the design manual. In order to achieve this, we propose a multi-function optimization method to extract parameters of a preliminary dimension- and distance-based process correlation model for interconnects. We propose an interconnect matching extraction method and suitable patterns useful for mismatch extraction. We have extracted matching guarantees while also considering the special case of edge matching using Monte Carlo analysis and field solvers. We have substantiated the experimentation with evaluations on an preliminary regression model and metric for mismatch.