Fast full-chip curvilinear ILT mask generation with machine learning technology

As an advanced form of optical proximity correction (OPC), inverse lithography technology (ILT), calculates the desired shapes on a photomask in free-form curvilinear types. It has become one of the key computational lithography solutions for advanced technology nodes, because it enables the most challenging features in IC designs and can produce mask output with better process latitude and CD control on wafer than using conventional OPC.

Although curvilinear ILT technology has taken root in limited cases, such as in hot spot repairment and in memory, it isn’t practical as a full-chip solution due to its long computation time and consistency issues. In this paper, we demonstrate how Calibre® pxOPCTM and Calibre SONR machine learning software from Siemens EDA provide a more efficient full-chip curvilinear ILT mask generation flow.

What you'll learn:

• How clustering-based techniques can accelerate full-chip curvilinear mask generation without sacrificing accuracy, addressing one of the main barriers in using ILT at scale.
• The role of Calibre tools (pxOPC, SONR, and MATE-SRAF) in improving process latitude, edge placement accuracy, and mask consistency through optimized curvilinear mask workflows.
• Benefits of integrating machine learning for feature clustering and pattern selection, which reduces computational load and enables full-chip ILT mask generation in less time.
• Performance comparisons between traditional ILT and this optimized curvilinear mask generation flow, showing equivalent lithographic fidelity with a more efficient runtime.

"By combining machine learning clustering, rule-based SRAF, and gradient-based optimization, the CFCC flow achieves ILT-level accuracy with significantly reduced runtime.."

- Yuansheng Ma, Author