Skip to content
Provided by ASME Logo The American Society of Mechanical Engineers

Track Keynotes

Ali Heydari

Organizer: Ali Heydari, Distinguished Data Center Technologist, Nvidia
Moderator: Gamal Rafai-Ahmad, AMD
Keynote Speaker: Ali Heydari, Distinguished Data Center Technologist, Nvidia

 

Abstract: Artificial intelligence and machine learning applications are about to permanently change design of data centers where liquid will be coming closer than ever as the common medium to cool the core of computational servers from GPU, CPU, Switch, and other components. Hybrid air and liquid cooling with direct to chip cooling design is going to be the low hanging fruit of choice for designers where liquid will be used to directly cool high heat density components while air will continue to cool other components. Design of Liquid plumbing, selection of cooling distribution units, selection of compatible wetted materials list and reliability/serviceability issues are some of the challenges that industry is striving to resolve as we see more data centers preparing to embrace liquid for cooling servers and other IT equipment. Utilizing CFD, FNM and Omniverse tools to create digital twins of high heat density data centers to address many challenges of high heat density data centers in the age of AI is presented.

Biography: Ali Heydari is Distinguished engineer and Data Center Technologist at Nvidia in charge of all data center cooling technology development at Nvidia. In this role, he is developing direct to chip liquid cooling technologies using cold plates, cooling distribution units and manifolds for cooling of Nvidia's high heat density data centers. Prior to Nvidia, he worked as senior director in charge of Rigetti’s Quantum Computers using the most futuristic technology in today’s data center compute. Accomplishments include, setting up the first Quantum Cloud Services enabling over the cloud access of the Quantum Computers. Prior to that he served as Senior Technical Director and Chief Data Center Architect at Baidu, the largest search engine and AI company in China. In this role, he was server and data center architect in charge of hardware and data center design, development and deployment in China’s largest data center search and AI company.

Formerly, he was Senior Hardware Engineer at Twitter where he was responsible for grounds up development of Twitter's data center ODM server development. Earlier, he was Senior Hardware Engineer at Facebook where he helped in developing Facebook’s original OCP server and data center products. Prior to that he worked at Sun Microsystems and spend about 10 years as Associate Professor of Mechanical Engineering at Sharif University of Technology in Iran. He received his B.S. in mechanical engineering from University of Illinois, Urbana, M.S., Ph.D. in mechanical engineering and M.A. in applied mathematics from University of California, Berkeley. He has over 25 issued patents in data center cooling technologies.

 

John Yundt

Organizer: Benjamin Leever
Moderator: Benjamin Leever
Keynote Speaker: John Yundt, SunRay Scientific

Abstract: Advancements in the development of Flexible Hybrid Electronics (FHE) materials and technologies continue to progress rapidly; nonetheless, high-volume manufacturing of complex FHE-based products with high reliability remains elusive. This issue becomes even more noteworthy especially in regards o printable wearable applications and technologies. Beyond flexible electronic devices that are inherently bendable, stretchable electronic devices are the next evolutionary developmental need for a wide array of real-world applications. Due to the nature of the biaxial strain induced during use, such stretchable devices are expected to be mechanically robust while maintaining electrical performance even under dynamic high tensile strains. Most of these stretchable electronic devices are hybrid in nature; being comprised of both soft and rigid electronic components. Hence, robust and reliable electrical interconnections between such soft and rigid components are necessary to ensure proper functionality of the device. Currently, interconnections between functional electrical systems and the wearable e-textile materials utilize traditional methods such as high-pressure Anisotropic Conductive Adhesives or Films (ACA/ACF), rigid solder connections, snap/cinch connections, and isotropic conductive adhesives (ICA). Each of these interconnection methods pose a variety of issues with the necessary material sets for wearables/e-textile materials. There are significant drawbacks related to mechanical and/or thermal damage during the connection process; poor mechanical adhesion; inability to connect standard-pitch functional devices; and lack of robustness when connecting to nonplanar surfaces seen in wearable device configurations. Additional challenges with these interconnect methods limit high-volume manufacturing of such products. SunRay Scientific Inc. will present recent results in the development of a novel anisotropic conductive adhesive, ZTACH® ACE, for high-volume surface mount assembly of electrical components, including those elements mounted/attached to flexible, stretchable membranes and textiles in wearable applications.

Prior research and development have demonstrated the technology behind ZTACH® ACE with superior performance in achieving environmentally stable and mechanically robust electrical connections. For improved manufacturability, this interconnect method allows for pressure-less assembly, low-temperature cure, excellent adhesion to a variety of substrates, and fine pitch reliability without sacrificing contact resistance or mechanical bond integrity. The ability of ZTACH® ACE to concurrently serve as its own underfill and edge encapsulant improves reliability especially with respect to the rigors of use in typical wearable applications on host substrates like textiles and TPU. It has been shown that ZTACH® ACE interconnects do not experience the same types of failures conventional competing technologies experience without the addition of underfill and/or encapsulation. In providing superior adhesion, low contact resistance, and mechanical robustness during electromechanical testing, ZTACH® ACE proves itself to be a reliable interconnect solution between stretchable to stretchable/flex/rigid materials with high electrical conductivity. With demonstrated scalability, these attributes enable easier and lower cost manufacturing of more flexible and robust applications through integration of SMD components directly onto e-textiles, without needing additional laminated protective layers. As a result, flexible wearable printed electronics can readily be incorporated into a wider range of cost-effective, reliable end use applications.

Beyond Flexible – Stretchable Electronic applications, ZTACH® ACE is suitable for semiconductor interconnect needs given high-yield finer pitch attributes. Successful developments have been achieved with demonstrations of IR FPAs, large-format area arrays, and RF applications. Continuing materials and processing maturation of this technology is expected to substantially address reliability and longevity requirements along with appreciable low-cost advantages most needed for semiconductor packaging and assembly.

Biography: John M. Yundt Vice President of Sales & Business Development, SunRay Scientific, Inc.
John is a graduate of Western Illinois University, with 28 years of experience in printed electronic materials and conductive coatings. Beginning his career at Acheson Industries (now part of Henkel), John became Business Development Manager, North America before leaving to start the Printed Electronics business at Spraylat Corporation. For the past decade John was Global Business Manager for PPG’s Electronic Materials business prior to joining SunRay Scientific in July of 2021.

 

Prof. Alberto Castellazzi

Organizer: Gilbert Moreno, NREL
Moderator: Emre Gurpinar, Sikorsky Aircraft
Keynote Speaker: Prof. Alberto Castellazzi, Kyoto University of Advance Science

Abstract: The electrification of pivotal elements of our societal infrastructure is increasing the demand for energy in electrical form. In turn, that is generating significant R&D momentum in efficient energy processing. Modern power conversion systems rely fundamentally on solid-state devices to shape the transfer of energy between source and load according to set performance targets. In the semiconductor arena, disruptive progress has been made with the introduction of wide-band-gap (WBG) technology, primarily silicon carbide (SiC) and gallium nitride (GaN). However, unleashing its full potential and justifying its higher costs in large volume deployment still require the development of bespoke packaging and circuit solutions. This talk will review learning done hitherto and discuss the way forward in the full exploitation of the superior features of WBG power devices.

Biography: Alberto Castellazzi is a Professor in the KUAS Engineering Department, where he leads research and teaching in advanced solid-state power processing, including the characterization and use of wide-band-gap (WBG) semiconductor devices (silicon carbide, SiC; gallium nitride, GaN), their packaging and thermal management, to yield disruptive joint progress in efficiency, power density and reliability of power converters. He holds a Laurea degree in Physics from the University of Milan, Italy, 1998, and a PhD degree in Electrical Engineering from the Munich University of Technology (TU Munich), Germany, 2004, and has been involved in power electronics research and development for over 20 years, with affiliation and working experience both in industry and academia, including SIEMENS Corporate Technology (Germany), ETH Zurich (Switzerland), ALSTOM Transport (France) and The University of Nottingham (UK). Alberto has published over 200 papers in specialist international journals and conference proceedings and has held a number of invited talks, tutorials, and seminars on the topic of WBG power devices and SiC-based electrical power conversion. He is active as both a reviewer and editor and is a member of the technical program committee of a number of international conferences on power electronics and power semiconductor devices.

 

Patrick McCluskey

Organizer: Patrick McCluskey, University of Maryland
Moderator: Pradeep Sharma, University of Houston
Keynote Speaker: Patrick McCluskey, University of Maryland

Abstract: Prof. Abhijit Dasgupta has been a leader in modeling the reliability of electronic packaging materials for four decades from the 1990s to the 2020s. During his career, he has made major advancements in constitutive property determination, multi-scale modeling, lead-free solder joint reliability, aerosol jet printing, and characterization of additively manufactured materials, including his seminal work on Energy Partitioning. His students have gone on to hold leadership positions in academia and industry, including Department Chair at a major research university and Director of Reliability at Microsoft. And he has been a major force in the success of the annual InterPACK conference, serving for many years on the organizing committee. In honor of Prof. Dasgupta, his former and current students have gathered to host a special mini symposium in the Reliability Track at this year's InterPACK conference. This keynote will highlight some of his major accomplishments in research, teaching, and service to this conference along with some anecdotes gleaned from working with him at the CALCE center for over 25 years. It will also discuss anticipated future developments in the reliability assessment of electronic materials that will grow from his foundational work.

Biography: Dr. Patrick McCluskey is a Professor of Mechanical Engineering at the University of Maryland, College Park, and the Department's Director of Undergraduate Studies. He has over 25 years of research experience in the areas of thermal management, reliability, and packaging of electronic systems for use in extreme temperature environments and power applications. Dr. McCluskey has co-authored three books, 5 US Patents, and nearly 200 peer-reviewed technical articles with over 3700 citations. He is an associate editor of the IEEE Transactions on Components, Packaging, and Manufacturing Technology, a member of the board of governors of the IEEE Electronic Packaging Society, a fellow and member of the Executive Council of IMAPS, and a member of ASME.

 

Krishna Gade

Organizer: Azeem Sarwar, General Motors
Moderator: Azeem Sarwar, General Motors
Keynote Speaker: Krishna Gade, Fiddler AI

Abstract: Artificial Intelligence is increasingly playing an integral role in determining our day-to-day experiences. Moreover, with the proliferation of AI-based solutions in areas such as hiring, lending, criminal justice, healthcare, and education, the resulting personal and professional implications of AI are far-reaching. The dominant role played by AI models in these domains has led to a growing concern regarding potential bias in these models, and a demand for model transparency and interpretability. Explainable AI has become a prerequisite for building trust and adoption of AI systems in high-stakes domains requiring reliability and safety such as healthcare and automated transportation, and critical industrial applications with significant economic implications such as predictive maintenance, exploration of natural resources, and climate change modeling.

In this talk, we will be talking about how Fiddler.AI is building an Explainable AI Platform that solves this problem by continuously monitoring AI algorithms for performance and bias issues and reports actionable insights with explanations to the entire organization. For more information, you can visit www.fiddler.ai or follow us on Twitter @fiddlerlabs.

Biography: For most of the last two decades, Krishna Gade spent time building scalable platforms at internet companies like Bing, Twitter, Pinterest, and Facebook to convert data into intelligent insights using big data, machine learning, and deep learning technologies. At Facebook, Krishna was leading the News Feed Ranking Platform that created the infrastructure for ranking content in News Feed and powered use-cases like Facebook Stories and recommendations like People You May Know, Groups You Should Join, etc. His team built Facebook's explainability features like Why am I seeing this? which helped bring much-needed algorithmic transparency and thereby accountability to the News Feed for both internal and external users. He is the co-founder and CEO of Fiddler AI. Fiddler AI is the first Explainable Monitoring solution for production AI systems, with the aim to build trust into AI by fully automating the monitoring and explainability of AI models in production.