This webinar is brought to you by
June 25 (Live Webinar) – 08:00 AM PST
08:00 – 08:10
08:15 – 08:35
Keynote: The Uncapped Potential of AI Chipsets
Presented by: Nicolas Sauvage, President & MD – TDK Ventures
View Full Details
08:40 – 09:00
The Cambrian Explosion in AI Silicon
Presented by: Karl Freund, Founder and Principal Analyst – Cambrian AI Research
View Full Details
09:05 – 09:25
AI Performance Innovation for Intel’s Current and Next-Gen AI Computing
Presented by: Huma Abidi, Senior Director of AI Products & Engineering – Intel
View Full Details
09:30 – 09:55
Heterogeneous IC Packaging for Advanced AI Applications
Presented by: Mike Kelly, VP, Advanced Package & Technology Integration – Amkor Technology, Inc.
View Full Details
10:00 – 10:20
Computing-in-Memory Circuit Design for Edge AI
Presented by: Shyh-Shyuan Sheu, Division Director, Electronic & Optoelectronic System Research Laboratories – ITRI
View Full Details
10:25 – 10:45
Backstage Pass: Reinventing Chips for AI
Presented by: Madhura Nataraju, Head of Products – Groq
View Full Details
10:50 – 10:55
President & MD – TDK Ventures
Nicolas Sauvage is Managing Director and Founder of TDK Ventures, the Corporate VC arm of TDK, with $200M AUM. He is responsible for TDK Ventures portfolio investment strategy and portfolio development. Nicolas is an alumnus of ISEN, London Business School, INSEAD and Stanford GSB.
It is hard to comprehend the exponential potential – and corresponding impact to our planet – the uncapped potential of AI Chipsets will bring. Yet, let’s try to comprehend it together.
TDK Ventures Inc. invests in startups to bolster innovation in materials science, energy/power and related areas typically underrepresented in venture capital portfolios. Established in 2019 as a wholly-owned subsidiary of TDK Corporation, the corporate venture company’s vision is to propel the digital and energy transformations of segments such as health and wellness, next-generation transportation, robotics and industrial, mixed reality and the wider IoT/IIoT markets. TDK Ventures will co-invest and support promising portfolio companies by providing technical expertise and access to global markets where TDK operates. Interested startups or investment partners may contact TDK Ventures: www.tdk-ventures.com or
Founder and Principal Analyst – Cambrian AI Research
Karl Freund is the founder and principal analyst at Cambrian-AI Research, where he helps refine and communicate the complex world of AI chips and platforms into easy-to-understand marketing materials. Previously, Karl worked as an analyst for Moor Insights & Strategy, and as an executive at AMD, IBM, Cray, and Hewlett-Packard. He holds a Masters degree in Computer Science from UNC Chapel Hill, and a BS in Applied Mathematics at Texas A&M University.
The world of AI demands ultra-fast chips in the data center, and efficient platforms for the edge, including autonomous vehicles. This presentation provides and overview of the major players (NVIDIA, Intel, AMD, Qualcomm, Amazon, Baidu, and Google) and unicorns (Cerebras, Graphcore, Tenstorrent, and others), to help the audience understand their strengths and weaknesses.
Cambrian-AI Research is an Industry Analyst firm dedicated to and solely focused on the AI and HPC Hardware market. Our services include advice and marketing assistance to help Semiconductor vendors, their investors, and technical audience to communicate and understand the explosion of new AI platforms that are shaping our future. www.cambrian-ai.com. Email at , +1-512-632-3634.
Senior Director of AI Products & Engineering – Intel
Huma Abidi is a Senior Director of AI Software Products & Engineering at Intel, She leads a globally diverse team of engineers and technologists responsible for delivering world class products that enable customers to create AI solutions.
Twice she has received the Intel Achievement Award – *Intel’s highest honor* and thrice awarded with the Intel Software Quality award, for delivering quality SW.
She is passionate about women’s education, supporting several organizations around the world for this cause, she was finalist VentureBeat 2019 for Women in AI award in Mentorship Category.
Huma holds a BS in pre-med and chemistry and an MS in computer science from the University of Massachusetts.
She has been delivering keynotes at several major conferences, recently at the O’Reilly AI conference in San Francisco, MIT PAN Arab conference in Lebanon, HPC Dell Conference in India, WEX Tech Talk in Portland, Maine etc.
Artificial Intelligence (AI) Deep Learning (DL) workloads require massive compute for training and Inference. However, End-To-End pipeline for AI workloads also require fast access to data, preprocessing, feature engineering, ETL, before AI Training. During Training, customers may also choose to quantization to lower precision. In addition, models may also require post-training quantization to accelerate inference. In this talk, we will present an overview of Intel’s OneAPI AI Analytics Toolkit. The toolkit provides Intel optimized libraries for AI Frameworks supported with Intel’s distribution of Python, Scikit-Learn, XGBoost, Modin, OneDAL, OneDNN, OneCCL, LPOT. The optimized libraries result in orders of magnitude performance improvement on Xeon® Ice Lake and next-gen Sapphire Rapids™ CPUs. We also plan to extend software acceleration with scalable abstraction for Intel’s general purpose GPU Xe architecture Ponte Vecchio.
Intel Corporation is an American multinational corporation and technology company headquartered in Santa Clara, California, in Silicon Valley. It is the world’s largest and highest-valued semiconductor chip manufacturer on the basis of revenue, and is the developer of the x86 series of microprocessors, the processors found in most personal computers. Intel ranked No. 46 in the 2018 Fortune 500 list of the largest United States corporations by total revenue. Intel is incorporated in Delaware.
VP, Advanced Package & Technology Integration – Amkor Technology, Inc.
Mike joined Amkor in 2005 and has led package developments for EMI shielding, thermally enhanced packages, sensors and high density MCM packages including 2.5D TSV and high-density fan-out (HDFO). He has worked in electronics and IC package design and manufacturing for 25 years, managing projects ranging from polyester flexible circuits to eutectic flip chip, IC package design and signal integrity. Mike has more than 40 patents in the field and holds master’s degrees in Mechanical and Chemical Engineering.
Massive data and data correlation techniques using artificial intelligence (AI) and machine learning are providing exciting breakthroughs in the areas of molecular protein research, faster-than-human obstacle recognition in vehicles, real-time Google Maps route recommendations, auto pilot systems on commercial airlines and even ride-sharing apps. This virtuous cycle of data creation from end devices feeds into large data centers with high-speed number crunching in a manner that was nearly unthinkable just a decade ago.
To meet large data demands Amkor has developed a high-performance packaging portfolio using key die and module fabrication technologies to combine a few or many die including logic, memory and more to realistically permit the integration of chiplets and high bandwidth memory (HBM) designs to provide heterogeneous packaging solutions.
You may know us for high-quality semiconductor packaging and test services. But do you know what inspires us to do what we do? An unswerving commitment to customer service, engineering expertise, and high volume manufacturing excellence. Helping solve the complex technological challenges facing the industry is what continually motivates us. We want our customers to focus their resources on semiconductor design and wafer fabrication while trusting us to be their reliable packaging technology innovator and test partner.
Amkor’s operational base includes production facilities, product development centers and sales and support offices located in key electronics manufacturing regions in Asia, Europe and the USA. As the original pioneers of the OSAT industry, Amkor Technology has helped define and advance the technology manufacturing landscape. Since 1968, we’ve delivered innovative packaging solutions with the service and capacity global customers rely on. We’re proud of what we’ve done and can’t wait to show you what’s next.
晶圆加工，并将我们视为其可靠的封装技术创新者与测试合作伙伴。Amkor 的经营场地包括位于亚洲、欧洲和美国主要电子制造地区的工厂设施、产品开发中心和销售及支持办公室。作为OSAT 行业的创新先驱者，Amkor Technology 一直帮助定义和优化技术制造现状。自 1968 年起，我们持续提供创新的封装解决方案，以及全球客户都能信赖的服务和功能。我们以此为豪，并迫不及待地向您展示美好的未来前景。
Division Director, Electronic & Optoelectronic System Research Laboratories (EOSL) – ITRI
Shyh-Shyuan Sheu received the Ph.D. degree from the Department of Electrical Engineering, National Central University, Chungli, Taiwan, in 2012.
He joined ITRI to develop the emerging semiconductor technology and circuit at 2003. Until now, he has published over 20 journals, 40 conference papers and 70 patents. Currently, he is the division director at the Chip Technology and Design Division, Electronics and Optoelectronic System Research Lab. (EOSL) of Industrial Technology Research Institute (ITRI) in Hsinchu, Taiwan. His research involves memory, emerging AI architecture and circuit(computing-in-memory), neuromorphic computing, 3DIC, display driver, and sensors.
- Deputy Division Director, Nanoelectronic Technology Division, Electronic and Optoelectronic System Laboratories, ITRI
- Assistant Professor, Department of Electrical Engineering, Yuan Ze University
- Deputy Research Director, Nanoelectronic Technology Division, Electronic and Optoelectronic System Laboratories, ITRI
- Manager, Integrated Circuit Design and System Application Department, Nanoelectronic Technology Division, Electronic and Optoelectronic System Laboratories, ITRI
- Research Manager, Integrated Circuit Design and System Application Department, Nanoelectronic Technology Division, Electronic and Optoelectronic System Laboratories, ITRI
- Research Deputy Manager, Integrated Circuit Design and System Application Department, Electronic and Optoelectronic System Laboratories, ITRI
AI with IoT (AIOT) application is widely appear in human’s life such as self-driving car, smart door bell. These application shows the AI is moving forward from cloud to edge device. However, the edge device for AI has the challenges of low to medium computing power, small form factor, long battery life time. These are all related to computing energy efficiency. The conventional computer architecture (Von-Neumann machine) is not efficient due to the memory bandwidth limit and data movement frequently. A new computing architecture computing-in-memory (CIM) is proposed to improve the computing efficiency. The CIM can directly compute (multiple and accumulate) in memory to avoid the data movement between the CPU and memory. However, CIM still have many challenges need to be overcome, such as linearity, precision and variation. In this talk, we proposed methodology to improve the precision and variations. It implemented to a 28nm 384kb 6T-SRAM Computation-in-Memory Macro with 8b Precision. This macro under 8bIN-8bWeights and 16 channel MAC operations can achieve 22.75TOPS/W for near full output-ratio (20b-OUT). This macro also achieved an FoM greater than 6× higher than previous SRAM-CIM work.
ITRI is a world-leading applied technology research institute with more than 6,000 outstanding employees. Its mission is to drive industrial development, create economic value, and enhance social well-being through technology R&D. Founded in 1973, it pioneered in IC development and started to nurture new tech ventures and deliver its R&D results to industries. ITRI has set up and incubated companies such as TSMC, UMC, Taiwan Mask Corp., Epistar Corp., Mirle Automation Corp., and Taiwan Biomaterial Co.
ITRI is dedicated to creating new value and identifying emerging demands for partners while facing global challenges and opportunities, such as urbanization, aging societies, new production and consumption patterns, climate change caused by global warming, and a post-COVID-19 era.
Head of Products – Groq
Madhura Nataraju is the Head of Products at Groq and is responsible for Groq’s hardware and software solutions. Prior to Groq she held a product leadership role at McAfee. She built her career in the semiconductor industry while working at Intel for over a decade in various engineering and business roles. She holds a PhD in Mechanical Engineering from the University of Wisconsin-Madison & Masters from Purdue University.
Every AI accelerator will talk about why their product is the best. To accomplish a leading status, we’ll look at what it takes to win as a chip maker. In this session we will overview the difficult choices facing the tech industry, the demands on compute, and ultimately why Groq’s architecture is so radically different.
Groq is an AI accelerator start-up that designed the Tensor Streaming Processor (TSP), branded as GroqChip™. Groq’s mission is to unleash the full potential of AI by replacing complexity with radical simplicity—allowing performance to skyrocket while effectively driving the cost of compute to zero. Groq approached its design with a software-first mindset, creating a simpler processor architecture and allowing software to control/manipulate the most complex allocations of compute resources.
View more of our events below