CASPA 2018 Spring Symposium: AI and Semiconductor Fusion - Part II Go Deep
Mar 3, 2018

AI and Semiconductor Fusion - Part II Go Deep

Artificial Intelligence (AI) has seen resurgence in recent years fueled by the advance in computational power, algorithms and large amount of data, which are propelled by the semiconductor chip technology. AI technology in turns creates new demand for semiconductor industry. The theme of the symposium will be  "AI and Semiconductor Fusion". This term was coined by CASPA in 2017 Annual Conference and become instantly popular among Semiconductor professionals and general tech audiences. We have historically high attendance in our annual conference. The 2018 spring symposium will go deeper on this theme with focus on the following areas:

(1). Deep learning applications and semiconductor solutions

(2). Enabling memory and computation technology for AI



Spring Symposium Agenda:

12:00 – 13:00 Registration and Networking

13:00 – 13:15 Welcome from CASPA President

13:15 – 15:00     Speeches:

                            FPGA Solution for AI, Vincent Hu, VP, Intel

Enhancing and Augmenting Human Perception and Health with Advanced Sensors and Artificial Intelligence, Achin Bhowmik, CTO, Starkey

Future Computing and Hardware, Dr. Jeffery Wesler, VP, IBM

Is It Time to Build Security in AI Chips?, Lan Jenson, CEO, Ada

15:00-15:15 Break and networking

15:15 – 16:30 Unleash Intelligence and Data Insights at Scale, Dr. Wei Li, VP, Intel

AI Trends and Impact on Memory system, Dr. Steven Woo, VP, Rambus

Enabling 10,000 Custom AI Chips at Scale with RISC-V, Dr. Yunsup Lee, CTO/Co-found SiFive

Sign-up for the event at:

Keynote Title:

Enhancing and Augmenting Human Perception and Health with Advanced Sensors and Artificial Intelligence


In the recent years, there has been an astounding pace of advances in the areas of sensing and artificial intelligence technologies. These breakthrough developments are increasingly enabling us to create devices and systems that can sense and understand the world around them. In this keynote, we will present a synopsis of the current state-of-the-art results in the field of augmenting the human sensation and perceptual processes with applications based on novel transduction devices and artificial intelligence technologies. The presentation will also highlight the emerging trends in these technologies as well as the associated business opportunities, especially in wearable devices that enhance human health.


Dr. Achin Bhowmik is the chief technology officer and executive vice president of engineering at Starkey, the largest hearing technology company in the US, a privately held medical devices business with more than 5,000 employees and operations in more than 100 countries worldwide. In this role, he is responsible for overseeing the company’s research and engineering efforts, towards redefining the medical wearable devices with advanced sensors and artificial intelligence technologies.

Prior to joining Starkey, Dr. Bhowmik was vice president and general manager of the Perceptual Computing Group at Intel Corporation. There, he was responsible for the R&D, engineering, operations, and businesses in the areas of 3D sensing and interactive computing, computer vision and artificial intelligence, autonomous robots and drones, and immersive virtual and merged reality devices. Previously, he served as the chief of staff of the Personal Computing Group, Intel’s largest business unit.

As an adjunct professor and guest lecturer, Dr. Bhowmik advises graduate research and teaches courses on human-computer interactions and perceptual computing technologies at the University of California, Berkeley, Stanford University, Liquid Crystal Institute of the Kent State University, Kyung Hee University, Seoul, and the Indian Institute of Technology, Gandhinagar.

Dr. Bhowmik was elected a Fellow of the Society for Information Display (SID). He serves on the board of advisors for the Fung Institute for Engineering Leadership at UC Berkeley, the executive board for SID, and the board of directors for OpenCV. He also serves on the board of directors and advisors for several technology startup companies. He received the Industrial Distinguished Leader Award from the Asia-Pacific Signal and Information Processing Association. He has over 200 publications, including two books and 34 issued patents.

Speaker: Jeffery Welser, Vice President & Lab Director -- IBM Research - Almaden

Abstract: Future Computing and Hardware

While the industry's traditional computing roadmap continues to progress, pushing into the exascale regime in the coming years, there are still many problems these systems are not well suited to solve. As more applications focus on understanding large, unstructured datasets or solving massive, multi-variable optimizations, there is a need for exploring new systems, chips and even potentially devices and materials to better handle these workloads. This talk will discuss some new directions being driven by the rise of AI and quantum computing, and the impact that may have on future semiconductor and system development.


Dr. Jeff Welser, Vice President and Lab Director, IBM Research – Almaden in San Jose, California oversees scientists and engineers performing exploratory and applied research. Home of the relational database and the world's first hard disk drive, the Almaden laboratory today continues its legacy of advancing data technology & analytics, specializing in areas including: Watson cognitive technologies, brain-inspired computing, augmented intelligence, data curation and platforms, computing as a service, fundamental science and nanotechnology in the areas of magnetics, spintronics, polymers and theory. Almaden's Accelerated Discovery Lab, focusing on data-centric systems, ad-hoc infrastructure and analytics, is part of the IBM Research THINKLab client innovation

experience, focused on providing industry solutions in areas including finance, retail, food safety, water management and healthcare.

In his prior position, Dr. Welser was the Director of Strategy and Program Development for the Accelerated Discovery Lab and simultaneously the Director of Almaden Services Research, managing a portfolio of research into improved business processes, software and technology to enable IBM's Global Business and Technology Services organizations.

Dr. Welser joined IBM Research in 1995, working on novel silicon devices at the T.J. Watson Research Center, and

served as adjunct professor at Columbia University, teaching semiconductor device physics at the same time. In

2000, he took an assignment in IBM’s Technology Group, and then joined IBM Microelectronics in 2001, as leader

of high-performance CMOS device design. In 2003, he was named Director of high-performance SOI and BEOL

technology development, in addition to continuing work as the IBM Management Committee Leader for the Sony,

Toshiba, and AMD development alliances. In late 2003, Welser returned to Research as the Director of Next

Generation Technology Components, looking at technology, hardware, and software components for future IT

systems. Starting in 2006, Welser was on assignment to the Semiconductor Research Corporation (SRC) as the

Director of the Nanoelectronics Research Initiative (NRI), a consortium of leading U.S. Semiconductor firms which

partners with NIST, NSF, and state governments to support university-based research on future nanoscale logic


Dr. Welser received his PhD in Electrical Engineering from Stanford University. Welser holds 21 US Patents and has

published over 75 technical papers and presentations. He is a member of the IBM Academy of Technology, IBM’s

Growth and Transformation Team, IBM’s Technology Team, an IEEE Fellow, a member of the American Physical

Society, Chairman of the Bay Area Science and Innovation Consortium and has served on numerous Federal agency

and Congressional panels on advanced semiconductor technology.

Vincent Hu, VP, Intel



Vincent Y. Hu is vice president in the Programmable Solutions Group (PSG) and leads the group’s Strategy Innovation and Planning office at Intel Corporation. He has overall responsibility for defining and driving group strategy to support PSG’s long-term revenue objectives. Hu also manages product device and intellectual property (IP) product planning, manages the office of the group’s chief technology officer, and serves as PSG’s diversity and inclusion champion. 

Before assuming his current role in 2017, Hu led PSG’s customer experience organization. He joined Intel in 2015 with the acquisition of Altera Corp. Hu began his Altera career in 2003, serving initially as vice president of technical services. Subsequently, his role grew to encompass product marketing and corporate marketing as well. He also served as, vice president of worldwide field applications engineering and technical services. In that role, he oversaw field and factory customer technical enablement resources and managed the development and delivery of technical collateral to support Altera’s device, software and IP portfolio.

Before joining Altera, Hu was president, chief operating officer and a co-founder of LinCom Wireless Inc., a startup developing wireless chipsets. Earlier in his career, he spent 11 years with LinCom Corp., culminating in his role as vice president of the Wireless Products Group. He started his career as an engineer at Hughes Aircraft Co.

Hu earned his bachelor’s degree in electrical engineering from Rice University, his master’s degree in electrical engineering from the University of Southern California and his MBA degree from the Anderson Graduate School of Management at the University of California, Los Angeles. He holds a patent for inventing a method for generating programmable digital sine waves.

Title: AI Trends and the Impact on Memory Systems



Recent advanced in AI and neural networks have driven the adoption of these technologies across the spectrum from data centers to mobile devices. Improved voice recognition, natural language processing, and image recognition are just a few examples of areas where AI and neural networks have greatly improved user interfaces, processing, and ultimately the utility of data itself.  As the world’s digital data continues to grow at unprecedented rates, the need for technologies that can make sense of it all is greater than ever before.  Companies are increasingly turning towards specialized silicon to improve the speed and power-efficiency of AI and neural network, and the demand for higher memory and interface performance are naturally following.  In this talk I’ll discuss some of the recent trends in AI and neural networks, and discuss the impact on memory systems for current and future silicon designed for these applications.

Speaker: Steven Woo, Vice President, Systems and Solutions and Distinguished Inventor, Office of the CTO, Rambus Inc.



Steven Woo is Vice President, Systems and Solutions and Distinguished Inventor at Rambus Inc, where he works in the Office of the CTO on technology and business development efforts across the company.  Since joining Rambus in 1996, Steve has worked in various roles leading architecture, technology development, and performance analysis efforts, and in marketing and product planning roles leading strategy, planning efforts, and customer programs.  Steve has over 30 patents in the areas of memory, memory systems, and neural networks.  


Steve received BS Engineering and Master of Engineering degrees from Harvey Mudd College, and MS and PhD degrees in Electrical Engineering while working in the Computer Systems Laboratory at Stanford University.

Best Regards,


Wei Li, VP, Intel

Title: Unleashing Intelligence and Data Insights at Scale

Abstract: The world’s most valuable currency is Data…and it is appreciating at an impressive rate. Every business is becoming a data business. AI and advanced data analytics are reshaping the enterprise with new discoveries, better customer experiences, and improved products and services – enabled by actionable insight. In this presentation, you will learn how Intel is driving a broad and holistic hardware & software approach to powering artificial intelligence & advanced analytics solutions and unleashing intelligent & scalable insights. Several real-world AI use cases enabled with the latest technologies and optimization advancements will be highlighted.



Wei Li is vice president in the Software and Services Group and general manager of Machine Learning and Translation at Intel, responsible for software systems including machine learning, binary translation, and emulation. His team works with industry and academia on the software ecosystem, and with Intel hardware team on CPU design. Li holds 11 U.S. patents and a Ph.D. in computer science from Cornell University. He completed the Executive Accelerator Program at the Stanford Graduate School of Business. He was an adjunct professor in computer science at Stanford University. He served as an Associate Editor of ACM Transactions on Programming Languages and Systems (TOPLAS).

Lan Jenson, CEO of Adaptable Security Corp

Is It Time to Build Security in AI Chips?


Lan Jenson is the CEO of Adaptable Security Corp (Ada for short), the only organization empowering people to know whom to trust in the Internet era.  Ada also solves the pain points of small and medium sized businesses in cybersecurity with affordable expertise.

Ms. Jenson has extensive policy experience in government, private and nonprofit sectors. As a certified information security professional, Ms. Jenson has won many company-wide recognition and awards for successfully leading risk management, policy and compliance programs in financial and technology companies such as Charles Schwab and Cisco Systems. Ms. Jenson also serves as the Communications Chair on the ISC2 Silicon Valley Chapter Board of Directors.

Title: Enabling 10,000 Custom AI Chips at Scale with RISC-V

Abstract: In a world where Moore's law and Dennard scaling is no longer going to save us, we need more purpose-built silicon, especially for applications such as AI where efficiency matters.  However, building custom chips from scratch today is very costly -- one needs to pay upfront for all IPs and assemble a world-class hardware/software team with a diverse set of expertise.  So, the million dollar question becomes: how do we enable mass customization at scale?  In this talk, I describe SiFive's approach to enable mass customization using RISC-V and open-source.

Bio: Yunsup Lee is SiFive’s Chief Technology Officer and co-founder. Yunsup received his PhD from UC Berkeley, where he co-designed the RISC-V ISA and the first RISC-V microprocessors with Andrew Waterman, and led the development of the Hwacha decoupled vector-fetch extension. Yunsup also holds an MS in Computer Science from UC Berkeley and a BS in Computer Science and Electrical Engineering from the Korea Advanced Institute of Science and Technology (KAIST).