Home > Tools > Custom IC Design > Cadence CMP Predictor

Cadence CMP Predictor 

Predict and optimize interconnect thickness and chip topography variability

Cadence CMP Predictor enhances design performance and yield through model-based CMP hotspot detection and CMP-aware RC extraction.

Product ImageCadence® CMP Predictor turns the uncertainty of manufacturing process variation into predictable impacts, and then minimizes these impacts during the design stage. Using models developed with Cadence CMP Process Optimizer, a tool that allows silicon calibration of semi-physical models and optimization of CMP material and process parameters such as pressure, polish time and overall CMP uniformity, Cadence CMP Predictor provides full-chip, multi-level interconnect thickness and topography predictions for copper electrochemical deposition (ECD) and copper/dielectric chemical-mechanical planarization (CMP) processes.

CMP-related hotspots, such as copper pooling, can have detrimental effects on chip yield. The conventional rules-based approach to hotspot detection fails to capture long-range and multi-level CMP effects. Cadence CMP Predictor uses a highly accurate model-based approach to finding potential hotspot areas. It also feeds the thickness and topography variation data into extraction tools, enabling better RC and timing analysis.

  • Accurately predicts multi-layer thickness and topography variability using a model-based approach developed with model calibration tool, Cadence CMP Process Optimizer
  • Identifies potential problem areas (hotspots) that affect yield
  • Minimizes or eliminates hotspots through integration with Cadence Chip Optimizer
  • Interfaces with Cadence QRC Extraction to identify timing-related problems (such as race conditions) and potentially reduce process guardbands