Efficient full-chip SRAF placement using machine learning for best accuracy and improved consistency

Wang, Shibing; S, Baron; Kachwala, Nishrin; Kallingal, Chidam; Sun, Dezheng; Shu, Vincent; Fong, Weichun; Li, Zero; Elsaid, Ahmad; Gao, Jinwei; Su, Jing; Ser, Jung-Hoon; Zhang, Quan; Chen, Been-Der; Howell, Rafael C.; Hsu, Stephen; Luo, Larry; Zou, Yi; Zhang, Gary; Lu, Yen-Wen; Cao, Yu

doi:10.1117/12.2299421

Cited by 8 publications

(7 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, today all EDA companies that offer OPC products also offer ILT products of some kind, some used for fixing hotspots like Synopsys, 19,20,98 some for model-based SRAF generation like ASML Brion. [99][100][101] ILT was extended to EUV by Synopsys [102][103][104][105] and ASML Brion started exploring using deep learning (DL) in ILT for SRAF generation. [99][100][101] Despite steady, continuing research and development across academia and industry through the decade and demonstration of the use of ILT to correct full-chip designs, ILT was still seen as an advanced method for use in critical hotspots, rather than as a technique to be applied to full-chip mask generation.…”

Section: History Of Inverse Lithographymentioning

confidence: 99%

“…[99][100][101] ILT was extended to EUV by Synopsys [102][103][104][105] and ASML Brion started exploring using deep learning (DL) in ILT for SRAF generation. [99][100][101] Despite steady, continuing research and development across academia and industry through the decade and demonstration of the use of ILT to correct full-chip designs, ILT was still seen as an advanced method for use in critical hotspots, rather than as a technique to be applied to full-chip mask generation. Excessive computational run-times continued to render full-chip ILT impractical in production settings.…”

Section: History Of Inverse Lithographymentioning

confidence: 99%

“…Inspired by many success stories of ML in a broad range of artificial intelligence applications, both industrial and academic researchers are now actively developing ML solutions for challenging problems in computational lithography, including ILT. [99][100][101]121,[138][139][140] One of the first papers applying DL to ILT is from ASML Brion. Their 2017 paper shows how they use their freeform ILT engine to train an ILT DL model using a convolutional neural network (CNN) (Fig.…”

Section: Applying Deep Learning To Iltmentioning

confidence: 99%

“…99 Once this DL ILT model is trained, it is used to create the SRAF initial-seeding map and also as the initial seeding of their ILT engine. Such DL-based SRAF generation has been demonstrated in both DRAM and logic full-chip applications 100,101 and FLASH memory full-chip application. 121 While most research work has focused on using DL to accelerate ILT, Dr. Peng Liu from Synopsys showed DL can create ILT solutions all by itself, plus all models-including neural network-based 3D mask, imaging, and resist models-required for ILT.…”

Section: Applying Deep Learning To Iltmentioning

confidence: 99%

See 3 more Smart Citations

Inverse lithography technology: 30 years from concept to practical, full-chip reality

Pang

2021

J. Micro/Nanopattern. Mats. Metro.

View full text Add to dashboard Cite

In lithography, optical proximity and process bias/effects need to be corrected to achieve the best wafer print. Efforts to correct for these effects started with a simple bias, adding a hammer head in line-ends to prevent line-end shortening. This first-generation correction was called rule-based optical proximity correction (OPC). Then, as chip feature sizes continued to shrink, OPC became more complicated and evolved to a model-based approach. Some extra patterns were added to masks, to improve the wafer process window, a measure of resilience to manufacturing variation. Around this time, the concept of inverse lithography technology (ILT), a mathematically rigorous inverse approach that determines the mask shapes that will produce the desired on-wafer results, was introduced. ILT has been explored and developed over the last three decades as the next generation of OPC, promising a solution to several challenges of advanced-node lithography, whether optical or extreme ultraviolet (EUV). Today, both OPC and ILT are part of an arsenal of lithography technologies called resolution enhancement technologies. Since OPC and ILT both involve computation, they are also considered as part of computational lithography. We explore the background and history of ILT and detail the significant milestones that have taken full-chip ILT from an academic concept to a practical production reality.

show abstract

Section: History Of Inverse Lithographymentioning

confidence: 99%

Section: History Of Inverse Lithographymentioning

confidence: 99%

Section: Applying Deep Learning To Iltmentioning

confidence: 99%

Section: Applying Deep Learning To Iltmentioning

confidence: 99%

See 2 more Smart Citations

Inverse lithography technology: 30 years from concept to practical, full-chip reality

Pang

2021

J. Micro/Nanopattern. Mats. Metro.

View full text Add to dashboard Cite

show abstract

“…To seek a way out of the OPC solution quality versus OPC solution TAT dilemma, the computational lithography community has turned to the ever-maturing machine learning technologies for help. Continuous attempts have led to some fruitful results in this area using deep convolution neural network (DCNN) architecture and bidirectional recurrent neural network 1 – 9 For machine learning OPC, the majority of the researches have been focused on designing some scheme that can measure the neighboring environment of a segment, then using the constructed feature vector from those measurements to predict the OPC amount of the segment through a trained neural network model 10 …”

Section: Introductionmentioning

confidence: 99%

Can optical proximity correction solution be learned? The learning limit and a general learning framework

Shi

Yan

et al. 2022

J. Micro/Nanopattern. Mats. Metro.

View full text Add to dashboard Cite

Background: Optical proximity correction (OPC) is an indispensable technology that has been propelling the advancement of computational lithography technology. To tightly control edge placement error (EPE) and maintain lithography process window, the demands on OPC computational resources and OPC turnaround time are growing rapidly with alarming acceleration. To tame the trend, machine learning technologies have been explored; however, an in-depth discussion on OPC solution learning limit is still lacking.Aim: We aim to present an in-depth discussion on OPC solution learning limit and propose a general machine learning OPC framework that can be extended to curvilinear mask OPC technology.Approach: In this study, we first investigate the machine learning OPC learning limit by examining noninverse lithography technology (non-ILT) OPC solution space characteristics inherited from edge segmentation and control point setting rules and then propose a general machine learning OPC framework that can take full advantage of deep convolution neural network (DCNN) learning capability while being able to preserve mask data high resolution.Results: With this machine learning OPC framework, we have achieved models with average absolute model error <1 nm for 14-nm metal layer. With single GPU, the average time for machine learning OPC models to produce results of 3840 nm × 3840 nm area is 8.74 ms for single channel input model and 12.65 ms for six channels input model. Conclusions:For non-ILT OPC solution, there is an intrinsic learning limit inherited from edge segmentation rules. Machine learning OPC models should be content with learning low order OPC solutions. This intrinsic learning limit of non-ILT OPC solution may diminish for ILT OPC solution when the constraint on degrees of freedom of OPC solution is lifted. The machine learning OPC framework we proposed is general and extendable to curvilinear OPC technology.

show abstract

AI Computational Lithography

Shi

Zhao

Chen

et al. 2020

2020 China Semiconductor Technology International Conference (CSTIC)

View full text Add to dashboard Cite

Efficient full-chip SRAF placement using machine learning for best accuracy and improved consistency

Cited by 8 publications

References 0 publications

Inverse lithography technology: 30 years from concept to practical, full-chip reality

Inverse lithography technology: 30 years from concept to practical, full-chip reality

Can optical proximity correction solution be learned? The learning limit and a general learning framework

AI Computational Lithography

Contact Info

Product

Resources

About