Lam Research is a global leader in wafer fab equipment and services since 1980. It is the world’s second largest semiconductor equipment manufacturer. Its main product lines are in deposition, etch and strip and wafer cleaning. In R&D, Lam Research spent US$1.2 billion last year.
Lam Research sees intelligent electronics as a market driver. It is addressing intelligent electronics challenges from a capital equipment perspective. It is now exploiting the learning from advanced 300mm applications to enable capability at 200mm. There are major trends in factory automation as well. More important, what does industry 4.0 mean for the semiconductor companies.
Lam Research’s equipment intelligence strategies for enhancing process performance, reproducibility and productivity are:
* Solutions that deliver reduced defectivity.
* Solutions that deliver improved installed base performance.
* Extension to 200mm technology.
Krishnan Shrinivasan, MD, Lam Research India, during the recently held IESA Vision Summit 2019, said that typically, PCs and mobiles were driving the growth, followed by automotive electronics. Increasingly, it has been driven by diverse, connected applications, such as IoT, cloud, AI/ML, VR/AR, robotics and medical. However, IoT, by itself, is not enough to drive the industry. It is the contribution of all the other stuff that is necessary to support IoT, in the network, data centers, and ML, etc.
Lam Research estimates that by 2020, there will be 30 billion connected devices. That will likely continue to increase, and move up to 75 billion connected devices by 2025. There is a significant growth in devices that may only need less than or equal to 28nm process capability.
There are things with sensors. There are also leading-edge applications, such as data centers and analytics. The greater, leading-edge devices are used at data centers. There are smart homes, connected cars, and wearables, as well as MEMS/sensors, MCUs/MPUs, smarter devices and PMICs, integration and SoCs, and advanced packaging.
The adjacent markets and specialist technologies being used by Lam Research are MEMS, CIS, power, BCD, 3D TSV, and photonics. Traditional CMOS markets and applications include logic, DRAM, 2D flash and 3D NAND.
Lam Research now have updated machines and products that have improved repeatability, defect reduction, throughput improvement, and reduction in the abatement of greenhouse gases. It is also repurposing the machines for the sensors and IoT worlds.
Role of industry 4.0
The world is changing with industry 4.0. The electronics industry is changing the world, and Lam Research’s technology powers the electronics industry. Technology today is progressing as a double exponential function of time. AI/ML has the ability to disrupt everything, including the semiconductor industry. It is driven by big data, deep learning algorithms, and specialized processors. Industry 4.0 will see the computerization and digitalization of all processes.
How can industry 4.0 enable semiconductors? Earlier, sensors and process control were looking backward. In the future, there would be equipment intelligence, to improve wafer-to-wafer (W2W) uniformity, load-to-load (L2L), and chamber-to-chamber (C2C) matchings, using sensors and control technologies, and the outcomes with ML, feed-forward process control and virtual fabrication.
Lam Research is building self-monitoring and self-tuning into its machines, as well as self-configuration. Predictive maintenance is a huge application as well. Automating the set-up of these machines and calibrating them are also something it can do via AI/ML technologies. Smart tool capabilities are going to be the accelerators in its fabs. Lam Research hopes that it can deliver even better results for customers.
Lam Research is likely to achieve larger process windows, lower costs for customers, higher yields, etc. It hopes to store the data and analyze it at the right time, for the right purpose. It is also doing advanced productivity solutions. It is building an ecosystem that allows Lam Research to drive the overall equipment efficiency and virtual processing. It is building all of this on top of a big data/ML platform with leading-edge automation, sensors and process control.
Data organization allows high-speed data collection and storage, and right data at the right time, in the right format, for ML. Advanced productivity solutions provide an ecosystem of software and services designed to drive improved overall equipment. In virtual processing, computational etch and virtual fabrication software enable fast learning and process development.
A big data ML platform enables an integrated platform, supporting fast algorithm development and customization. By automation, robotics and platform design increase productivity and enable lights out operation. Sensors and process control in situ and integrated sensors monitor system and wafer status during process at a high sensitivity and data rate.
By Aanchal Ghatak & Pradeep Chakraborty