Method verifying process proximity correction using machine learning, and semiconductor manufacturing method using same
AssigneeSamsung Electronics Co., Ltd.
InventorsSeorim Moon, Bonhyun Gu, Sooyong Lee
Life of this patent
Estimating the term…
Abstract
A method of manufacturing a semiconductor chip includes; generating a layout pattern, performing Process Proximity Correction (PPC) on the layout pattern to generate a PPC layout pattern, wherein the performing of PPC includes verifying the PPC layout pattern using machine learning, performing Optical Proximity Correction (OPC) on the PPC layout pattern to generate an OPC layout pattern, manufacturing a mask using the OPC layout pattern, and manufacturing a semiconductor chip using the mask.
Claims20
A method of manufacturing a semiconductor chip, the method comprising:
- generating a layout pattern;
- performing Process Proximity Correction (PPC) on the layout pattern to generate a PPC layout pattern;
- verifying the PPC layout pattern, wherein verifying the PPC layout pattern comprises calculating a metric using machine learning and comparing the calculated metric to a stored metric;
- performing Optical Proximity Correction (OPC) on the PPC layout pattern to generate an OPC layout pattern;
- manufacturing a mask using the OPC layout pattern;
- and
- manufacturing a semiconductor chip using the mask.
A method of manufacturing a semiconductor chip, comprising:
- performing Process Proximity Correction (PPC) on a layout using a first machine learning model to generate a PPC layout pattern;
- and
- verifying the PPC layout pattern,
- wherein the verifying of the PPC layout pattern includes:
- calculating an After Cleaning Inspection (ACI) Critical Dimension (CD) (ACI-CD) using a second machine learning model;
- and
- detecting a weak pattern in the PPC layout pattern in relation to at least one of vertical structure susceptibility, an After Development Inspection (ADI) CD (ADI-CD) variation, and etch loading susceptibility.
A computing device comprising:
- a memory device configured to store a Process Proximity Correction Rule Check (PRC) tool;
- and
- at least one processor configured to execute the PRC tool,
- wherein the PRC tool is configured to calculate an After Cleaning Inspection (ACI) Critical Dimension (CD) (ACI-CD) using machine learning, and detect a weak pattern in a Process Proximity Correction (PPC) layout pattern using the ACI-CD.
Dependencies are inferred from the claim text.