Semiconductor enabling VR/AR as the new dimension of human connection

Yan, Ran
Business Unit Manager
GLOBALFOUNDRIES

Abstract
As the world Digitalization is rapidly accelerating and providing real benefit to each one of us, a spate of emerging technologies—especially, artificial intelligence (AI), virtual reality (VR), and augmented reality (AR) —are gaining prominence across industries. These technologies have their unique capability to reduce the distance between people and deliver fully immersive experiences in all kinds of environments. The power of VR/AR combined with AI also allows machines to operate at the cognitive level as humans and allows us to interact naturally with machines. Such emerging technologies bring huge opportunities to semiconductor, but also new challenges that need our attentions: see-through near eye microdisplay, real-time image sensing & processing, low-weight low-power for round the day usage. Those requirements are not only challenging for design house, but also for semiconductor technology. Instead of chasing down the advance node, the industry needs to collaborate vertically and find the right trade-off between speed, power and cost. In GlobalFoundries, we are taking up on this challenge with our partners to enable the next-generation VR/AR products based on our unique solutions. GlobalFoundries® (GF®) Microdisplay solutions are optimized to improve process speed and reduce leakage while enabling enhanced pixel driver functionality.

It significantly speeds up VR/AR applications to support real-time data analysis and edge computing. Our platform is compatible with multiple display technologies, like LCOS, and microLED. We are aiming to enable our technology down to 22nm to reach 4K resolution within 2 inches diagonal MicroDisplay. There are still more challenges to overcome before widespread consumer VR/AR applications. However, with our effort to advance foundry technology and collaboration with industry partners, in the not-too-distant future, we will see AR everywhere in our life and connect us in real-time without any “distance”.

Biography
After 10 years working in semiconductor and GLOBALFOUNDRIES® (GF®), I am so proud to be part of this vital industry and hold my exciting position as business line manager for Human-Machine-Interaction (HMI) products. My vision is to reshape HMI technologies all over the world and enable the AR/VR foundry solution with a special focus on MicroDisplay and image sensors. In GF®, we know we cannot do it alone. That is why I am glad to have our industry partners, research institutes, and government bodies support us, especial in Europe. In addition, we must have human needs in our hearts and do not forget about the minorities. Therefore, I am also a Diversity & Inclusion Partner in our German site to build better technology, a better workplace, and a better society. I hold an EMBA from ESCP Business school and a Ph.D. in Microelectronic Engineering from the National University of Ireland, Cork.

How Edge Computing Enables Predictive Valve Maintenance in the Semiconductor Industry

Klingstedt, Hans
Senior Project Manager
Smart Systems Hub

Abstract
In the presented use case, the goal was to replace the monitoring of production-critical ultra-pure water valves at the Dresden site of challenge owner GLOBALFOUNDRIES with a suitable AI-based sensor solution. Sensor-based monitoring of valves ensures predictive maintenance and uninterrupted production, not just in chip manufacturing. Defects in valves were previously unpredictable at challenge owner Globalfoundries - a U.S. semiconductor manufacturer with over 16,000 employees worldwide and the largest and most modern semiconductor plant in Europe.The scalable edge computing solution, developed jointly with Coderitter, Globalfoundries, Infineon, Sensry, T-Systems and a hub team, is based on special sensors that provide acoustic data. Attached to the valve a small multisensorplatform as a smart sensor edge device enables the fusion, analyzation and classification using machine learning algorithms. The solution also includes the forwarding of the data to the cloud and the clear presentation in dashboards.On the one hand, the case is highly relevant in the context of the worldwide lack of semiconductor chips. Creating “virtual” capacity by using AI-based predictive Maintenance solution is promising action not only for production plants of Globalfoundries but the whole industry. Finally, we will look at how solutions can be developed for companies at different stages of technology and market maturity and how this helps European industry from startups to large companies.

Biography
Career start as assistant to the board of directors at an automotive supplier group.Support of projects in supply chain and supplier management of C and E-Class series.2016 as in-house consultant, design of digital transformation with introduction of PLM and SAP systems.From 2020 onwards, specialisation in innovation management and support of co-innovation formats as well as projects in digitalisation consulting in the Smart Systems Hub.Developing innovative IoT testbeds and MVPs by guiding different project partners as well as cross-sector technology experts (industry, SMEs, start-ups) through an innovation process.Focus lies fast integration of IoT technologies and AI to solve a problem and align them in such a way that companies succeed in process improvements and develop new business models.

Engineered Substrates and Materials for 5G

Roda Neve, Cesar
R&D Program Manager
SOITEC

Abstract
information coming soon

Biography
Cesar Roda Neve was born in Madrid, Spain, in 1975. He received the Msc. Engineer degree from the ICAI Universidad Pontificia de Comillas, Madrid, Spain, in 2000. In 2012, he received the Ph.D. degree in engineering sciences from the Université catholique de Louvain (UCL), Belgium. From 2004 to 2006, he was with the Electronics Department of the University Carlos III of Madrid, Spain, where he worked on ROF links and optoelectronic devices. From 2006 to 2012, he joined the Microwave Laboratory at the Université catholique de Louvain (UCL), Belgium, where he worked on the characterization and application of Si-based substrates for RF integration, in particular the use of HR-Si, HRSOI, and trap-rich HR-SOI substrates, non-linearities and parasitic effects. From 2013 to 2016, he was with the 3D and Optical Technology group at IMEC, where he worked on signal integrity, power delivery networks and RF modeling with special attention to 3D stacking and packaging. From 2016 to 2020, he worked at M3Systems Belgium as project manager for GPS, interferences and satellite related projects. In 2021 he joined SOITEC Belgium as R&D Program Manager. His research interest are new applications for SOI substrates for RF, with focus in 5G and 6G communications.

Sustainable AI: measuring and reducing the carbon footprint of deep learning model development and inference

Kelleher, John
Academice Leader ICE Research Institute
Technological University Dublin

Abstract
Artificial Intelligence (AI) has become a pervasive technology in modern societies. Naturally this has resulted in questions being raised regarding the ethical use of AI. However, a relatively under-studied aspect of modern AI is the relationship between AI and the environment. Used correctly AI has the potential to help our societies become move environmentally sustainable. At the same time modern AI, and in particular large Deep Learning models trained with powerful computers using massive datasets, have a direct environmental cost. In this talk I will discuss the environmental cost of modern AI practices and describe some of the ongoing research that is attempting to make AI more environmentally sustainable.

Biography
John is a Professor of Computer Science at Technological University Dublin. He is the Academic Leader of the Information, Communication and Entertainment (ICE) research institute, and a co-Principal Investigator at the Science Foundation Ireland ADAPT research centre, and a co-Principal Investigator at the SFI centre for PhD training in digitally enhanced reality (D-REAL). John has over 25 years of research experience in Artificial Intelligence, with a focus on the topics of natural language processing and machine learning. John has authored three books: Fundamentals of Machine Learning for Predictive Data Analytics (2020, MIT Press), Deep Learning (2019, MIT Press), and Data Science (2018, MIT Press). John's lab carries out research on natural language processing, machine learning for health and alsoo on the carbon footprint of deep learning. John's presentation at SEMICON will be on this last topic, the environmental impact of artificial intelligence.