Choose language
Back

Advanced-Process PDK Development

Download

Empowering Advanced Process from R&D to Mass Production with Silicon-Data Accuracy

As process nodes cross the 3nm, GAA structures replace FinFET as the mainstream, and device 3D complexity rises exponentially. Even minute shifts in process parameters — such as fin height and gate pitch — erode timing margins and power budgets through the amplifying effects of parasitic capacitance and resistance. The accuracy of PEX (Parasitic Extraction) PDK parasitic modeling and EMIR rule check becomes decisive factors in the successful deployment of a new process. Using Silicon-Data-level modeling accuracy as its benchmark, Phlexing delivers a complete PDK development solution spanning PEX development, EMIR rule check, and tape-out preparation.

Core Highlights

  • Silicon-Data-Level Accuracy Benchmark

    Based on the FDM (Finite Difference Method), it builds a high-fidelity 3D process model using foundry measured silicon data as the benchmark, serving as the Golden standard for Rule Deck extractor accuracy calibration and eliminating systematic deviation.
  • Three-Tier Precision Synergy

    From Golden modeling to device-level field solving and finally full-chip rule verification, it covers the entire PDK parasitic development lifecycle end to end.
  • Multi-Corner Parallel Verification

    Equipped with MCPE (Multi-Corner Parallel Extraction), it covers all RC corners in a single pass, improving RC extraction efficiency by over 35%.
  • DTCO Toolchain Support

    Built on RC parasitic extraction, it provides unified data support across process R&D, PDK development, and design verification, accelerating the process iteration cycle.

Key Features

01

High-Fidelity 3D Process Model Establishment
Based on FDM solution of Maxwell's equations, it integrates process-effect modules such as BEOL, CMP, and Etch-Loading to build a high-fidelity 3D process model benchmarked against foundry Silicon Data. It supports Cco/Cf modeling and interconnect model establishment with high SPICE model fidelity, serving as the Golden standard for Rule Deck extractor accuracy calibration.

02

Device-Level High-Accuracy Parasitic Extraction
Based on the random-walk method for efficient solving of Maxwell's partial differential equations, it achieves 0.1nm capacitance solving accuracy for standard cells, SRAM, and block critical paths, covering Planar MOSFET, FinFET, GAA, and CFET across all process nodes. Calibrated against leading foundries' Golden data, it has received accuracy certifications from multiple foundries, providing a precise reference for Rule Deck development and Foundation IP optimization.

03

Full-Chip Rule Deck Accuracy Verification
Verifies full-chip PDK PEX Rule Deck accuracy; MCPE multi-corner parallel extraction covers all RC corners in a single pass.

04

Full-Flow EM Rule Development
Generates the tool's Tech File from EM rules, produces EM QA pattern GDS files, builds a QA flow for comparison validation, and ultimately delivers the foundry-release Tech File.

05

Pre-Tape-Out PDK Signoff Confirmation
Based on a three-layer closed-loop verification comprising Rule Deck results, Golden data, and silicon measurements, it completes integrity and accuracy confirmation of the PDK parasitic model before tape-out, reducing tape-out risk.

Applications

FinFET/GAA Advanced-Process PDK Development
A complete toolchain from device modeling to Rule Deck verification, covering 14nm and below process nodes.
Foundry PEX Accuracy Verification & Certification
Golden-standard accuracy comparison ensures Rule Deck consistency with silicon data, supporting PDK release.
Fabless PDK Adaptation & Optimization
Optimizes parasitic parameter configuration for specific design requirements, completing PDK accuracy confirmation and signoff preparation.
FinFET/GAA Advanced-Process PDK Development
A complete toolchain from device modeling to Rule Deck verification, covering 14nm and below process nodes.
Foundry PEX Accuracy Verification & Certification
Golden-standard accuracy comparison ensures Rule Deck consistency with silicon data, supporting PDK release.
Fabless PDK Adaptation & Optimization
Optimizes parasitic parameter configuration for specific design requirements, completing PDK accuracy confirmation and signoff preparation.

Anchored in Advanced-Node Signoff
Empowering a τ-Aware EDA Framework

Please provide your contact information

By providing your information, you agree to our processing of your personal data in order to
carry out the activity and related communications. For details, please refer to the Privacy Policy.

Submit