Machine learning models have been applied to a wide range of computational lithography applications since around 2010. They provide higher modeling capability, so their application allows modeling of higher accuracy. Many applications which are computationally expensive can take advantage of machine learning models, since a well trained model provides a quick estimation of outcome. This tutorial reviews a number of such computational lithography applications that have been using machine learning models. They include mask optimization with OPC (optical proximity correction) and EPC (etch proximity correction), assist features insertion and their printability check, lithography modeling with optical model and resist model, test patterns, and hotspot detection and correction.

Gradient Boosted Tree is a powerful machine learning method that supports both classification and regression, and is widely used in fields requiring high-precision prediction, particularly for various types of tabular data sets. Owing to the recent increase in data size, the number of attributes, and the demand for frequent model updates, a fast and efficient training is required. FPGA is suitable for acceleration with power efficiency because it can realize a domain specific hardware architecture; however it is necessary to flexibly support many hyper-parameters to adapt to various dataset sizes, dataset properties, and system limitations such as memory capacity and logic capacity. We introduce a fully pipelined hardware implementation of Gradient Boosted Tree training and a design framework that enables a versatile hardware system description with high performance and flexibility to realize highly parameterized machine learning models. Experimental results show that our FPGA implementation achieves a 11- to 33-times faster performance and more than 300-times higher power efficiency than a state-of-the-art GPU accelerated software implementation.

This paper presents a case study of FPGA implementation for electrical parts counting and orientation recognition method based on industrial vision system. Since parts counting and orientation recognition can be achieved using a trained neural network (NN), the paper studied its efficient implementation using FPGA. Contributions include weights binarization, activation function approximation and HW architecture design for NN recognition. Experimental results revealed that the proposed implementation achieved faster speed and lower hardware resource usage.

This paper reports a supplemental process design kit (PDK) for ASAP7 PDK using Synopsys design flow. ASAP7 is a PDK for “predictable” 7-nm FinFET technology node. ASAP7 PDK is useful for academical and educational purpose, however it only supports Cadence platform for Place and Route. A supplemental PDK is designed for ASAP7 to use Synopsys platform for Place and Route. This PDK is opened at the author's GitHub site for both acamemical and educational usage.

Ising machines have recently attracted much attention because they are expected to solve combinatorial optimization problems efficiently. We focus on an Ising machine whose algorithm is based on parallel tempering (PT), and experimentally evaluate the performance of the Ising machine for MIN-CUT problems. Experimental results show that the Ising machine outperforms a famous graph partitioning solver in terms of the quality of solution and the time-to-target-solution.

Drone is one of the promising vehicles that have exhibited the potential to reduce the cost and time in the field of logistics. However, due to the limitation of battery capacities, the flight time remains short. Therefore, energy consumption is one of the most critical concerns in drone delivery services. In order to reduce the energy consumption, drone generally needs to fly to the destination in as short a time as possible. For delivery services, the drone has loads to deliver and is exposed to weather effects such as windy conditions. This paper studies a routing problem for energy minimization of delivery drones under the assumption of windy conditions. This paper formally defines Energy Minimizing Vehicle Routing Problem (EMVRP) under windy conditions. Experimental scenarios with different wind velocities and the number of customers have been simulated, and demonstrate a comparison of the metrics in the energy consumption and the flight distance.