Table II from A Unified Framework for Layout Pattern Analysis with Deep Causal Estimation | Semantic Scholar (2024)

A novel layout-aware diagnosis-based layout pattern analysis framework is proposed to identify the root cause efficiently and outperforms a commercial tool with higher accuracies and around $\times 8.4$ speedup on average.

This article develops a new end-to-end hotspot detection flow where layout feature embedding and hotspots detection are jointly performed and an attention mechanism-based deep convolutional neural network is exploited as the backbone to learn embeddings for layout features and classify the hotspots simultaneously.

RCANet is developed, an end-to-end unsupervised learning-based RCA framework, which analyses diagnosis reports of failing dies within a wafer and identifies both layout-aware and cell-internal root causes efficiently.

This paper proposes several quantization algorithms specifically designed for a classic neural network based hotspot detector while taking into account the feature distribution of the dataset used, and shows that they can achieve competitive results compared with full-precision models while significantly reducing inference runtime at the same time.

This paper will survey the recent pace of progress on advanced methodologies for wafer failure pattern recognition, especially for mixed-type one, and hopes this literature review can highlight the future directions and promote the advancement of the waferFailure pattern recognition.

This paper will survey the recent pace of progress on advanced parameter auto-tuning flows of physical synthesis tools to reduce human cost and tool evaluation cost and hopes this survey can enlighten the future development of parameterAuto- Tuning methodologies.

A deep learning framework for high performance and large scale hotspot detection is developed and a biased learning algorithm is proposed to train the convolutional neural network to further improve detection accuracy with small false alarm penalties.

This work proposes the use of the area under the ROC curve (AUC), which provides a more holistic measure for im-balanced datasets compared with the previous methods, and proposes the surrogate loss functions for direct AUC maximization as a substitute for the conventional crossentropy loss.

A new end-to-end framework that can detect multiple hotspots in a large region at a time and promise a better hotspot detection performance is developed and Experimental results show that this framework enables a significant speed improvement over existing methods with higher accuracy and fewer false alarms.

By clustering images of the layout locations that correspond to diagnosed sites for a statistically large number of IC failures, LASIC uncovers the common layout features and is found to be effective.

A method that uses diagnosis to identify layout features that do not yield as expected and clustering techniques are applied to layout snippets of diagnosis-implicated regions from (ideally) a statistically-significant number of IC failures for identifying feature commonalties.

This work focuses on significantly increasing the value of test data and the yield learning rate, which is a critical factor in the success of an IC in the market place.

Surprising empirical results are reported that simple Siamese networks can learn meaningful representations even using none of the following: (i) negative sample pairs, (ii) large batches, (iii) momentum encoders.

Volume diagnosis data mining for root cause identification based on statistical methods is used to reduce turnaround time and cost to speed up the process of systematic defect identification.

This paper details the principles that drove the implementation of PyTorch and how they are reflected in its architecture, and explains how the careful and pragmatic implementation of the key components of its runtime enables them to work together to achieve compelling performance.

...

...