Discover Run-to-Run software is a feature-rich application that enables advanced supervisory process control, surpassing traditional statistical methods. It ensures precise process goals. For example, achieving desired thickness is possible through tailored adjustments of machine parameters like deposition duration or pressure settings. Designed for high-mix production, it allows seamless transfer of control strategies across factories for operational flexibility. Proven globally, it operates thousands of tools across hundreds of processes continuously. With its adaptable, framework-based APC solution, Discover Run-to-Run software enables rapid factory-wide deployment, delivering significant and immediate return on investment while enhancing consistency, productivity and efficiency.

Discover Yield software helps reduce the time it takes to find root causes of yield excursions and enables manufacturers to broaden their analysis scope for better yield management.

The software includes data acquisition and integration, a development environment through workflow and scripting, and specialized analysis algorithms to identify domain-specific issues, such as data mining, spatial anomalies, wafer processing sequence problems, commonality of effects, system and random yield loss, chamber matching, tool events, traceability and more. In addition, Discover Yield software provides distinct tools – principal components and multivariate analysis of variance (MANOVA) – that allow systematic, simultaneous examination of multiple variables. Discover Yield software’s database architecture for genealogy uniquely supports value chain predictive analytics.

Improving baseline yield with analytics

Discover AI software is a disruptive technology that reshapes the way we approach manufacturing line control and yield analysis. It leverages the latest in machine learning and artificial intelligence to identify actionable improvements and connections in data that a human could not. The software’s optimization engine adds a powerful machine learning analytical method to an already sophisticated toolbox of process improvement applications.

By analyzing any set of conceivable inputs and outputs, Discover AI software identifies relationships and interactions that lead to positive operational changes. Easily understood visuals empower even casual users to understand what is important before making adjustments.

Discover AI software evaluates multiple models and suggests the optimal one for the data. Using machine learning, it examines interactions between in-process inputs and results for any semiconductor product family and recommends changes to in-process targets. Licensed as a service, it enables users to leverage an on-demand infrastructure to apply easily understood workflows for complex analytics without the overhead.

An integrated metrology standard, the IMPULSE+ system offers high sensitivity and accuracy to CMP process excursions and enables device makers to establish APC control with high-accuracy feedback. The IMPULSE+ system works in conjunction with the Atlas® platform, facilitating cross-module process optimization and comprehensive fab-wide process control.

Based on a common optical design derived from our stand-alone Atlas® platform, the IMPULSE+ system shares superior deep ultraviolet (DUV) optics. This unique ecosystem can be leveraged to improve signal to noise ratio by applying spectra feedforward from the Atlas platform, further enhanced by AI-driven machine analytics. This combination affords CMP process engineers a potent arsenal of capability to manage excursions and drive process improvement (Cpk.).

The IMPULSE+ system is widely adopted across key steps in DRAM, 3D-NAND, CMOS image sensor and foundry/logic device manufacturing. challenging to measure.

As wafer uniformity requirements tighten and vertical scaling increases, the need for CMP processing grows across market segments. Logic introduces new transistor designs and materials, DRAM incorporates new materials and process steps to planarize, and 3D NAND adds more decks, stacks and tiers.

The IMPULSE V system, available as an integrated or standalone platform, enhances film measurement fidelity and productivity for the CMP process module. With deep ultraviolet (DUV) optics and AI-driven machine analytics derived from our Atlas^® platform, it works in conjunction with the Atlas^® platform to provide CMP process engineers the capability to manage excursions and drive process improvement (C_pk.).

Advanced wavefront technology suppresses previous layer noise in complex structures. It provides feedback to improve long term repeatability. Designed for higher sampling, in-die/on-device and wafer edge measurements, the system offers improved reliability, higher throughput and over 2X precision improvement compared to the previous generation. Onboard AI-driven machine learning uses a signal-to-noise ratio for faster time-to-solution, addressing layers that were previously challenging to measure.

The Element S system is designed for the specialty market, specifically the smaller 100mm to 200mm wafer sizes used in the rapidly growing Si/SiC/GaN power device market. Building on the success of the Element Fourier transform infrared (FTIR) system, the Element S system can accurately measure thickness and free carrier concentrations in epitaxial layers up to five layers. This capability is essential for characterizing the thick epitaxial films required for next generation SiC power devices, which need thick epi layers for higher standoff voltage.

With its small spot size, the Element S system can conduct measurements up to the very edge of the wafer, maximizing die yield and enhancing productivity as power device customers transition to 200mm wafers. With improved sensitivity, an advanced multi-layer epi algorithm, and leading-edge multiple IR peak modeling, the Element S system provides critical and versatile capabilities for wafer and device makers in epi and film composition monitoring.