Keynotes

Speakers

Prasad Chaparala

Prasad Chaparala is the Director of Reliability Engineering at Amazon Lab126 in Sunnyvale, California. He is responsible for reliability engineering of a broad range of consumer electronic devices such as Echo smart speakers, Kindle e-readers, Fire tablets, and Fire TV products. Prior to this, he was the Vice President of Product and Reliability Engineering at Alta Devices from 2010 to 2014. Before joining Alta Devices, he was with National Semiconductor for 14 years in various process and reliability engineering roles. He received a Ph.D in Reliability Engineering from the University of Maryland, College Park. He has contributed to more than 50 publications in international journals and conference proceedings and holds 18 US patents. He is a recipient of three Best Paper awards at IEEE International Reliability Physics Symposium (IRPS). Additionally, he was served as the General Chair for the 2014 IRPS and was a member of the Board of Directors for IRPS.

Keynote Abstract

Building Reliable Products Guided by Customer Obsession

With the rapid proliferation of consumer IoT devices that are embedded into everyday life, building both affordable and reliable hardware that continues to meet the highest customer expectations is of paramount importance. This is particularly challenging for new applications where customer-use conditions can vary broadly and can be unpredictable. Unlike other established industries such as semiconductor, automotive or aerospace where widely accepted reliability standards exists, there are no industry reliability standards for consumer electronics devices. In-depth understanding of customer usage environments, patterns, and expectations is critical in deriving appropriate system-level reliability specifications and test methods in order to build reliable devices that surpass customer expectations. This talk will provide an overview of how system-level reliability requirements for devices such as the Amazon Echo and Fire TV products are defined by working backwards from customer needs. The talk will cover various advanced engineering approaches in defining reliability specifications and test methods through user surveys, statistical analysis and machine learning techniques and customer feedback.

Shimeng Yu

Shimeng Yu is an associate professor of electrical and computer engineering at the Georgia Institute of Technology in Atlanta, Georgia. He received the B.S. degree in microelectronics from Peking University, Beijing, China in 2009, and the M.S. degree and Ph.D. degree in electrical engineering from Stanford University, Stanford, California, in 2011 and in 2013, respectively. From 2013 to 2018, he was an assistant professor of electrical and computer engineering at Arizona State University, Tempe, Arizona.

Keynote Abstract

Emerging Non-Volatile Memory’s Applications in Neuro-Inspired Computing and Hardware Security

Emerging non-volatile memory (eNVMs) technologies have made significant advances in the past decade as storage class memory and embedded memory with extensive industrial research and development. This presentation will survey the recent progresses of using eNVMs for new applications beyond data storage in the era of artificial intelligence (AI) and Internet of Things (IoT), in particular for 1) neuro-inspired computing and 2) hardware security. Firstly, I will introduce the individual components of the deep neural network hardware – the eNVM based synaptic devices and neuronal devices. Secondly, I will discuss the crossbar array architecture that embeds the computation into memory array, namely compute-in-memory approach, and show the array-level and chip-level demonstration results. Then, I will introduce the “NeuroSim” framework, a device-circuit-algorithm co-design simulator that benchmarks the non-ideal effects of eNVMs on the machine learning accelerator performance. Lastly, I will introduce how to leverage the eNVM’s variability as physical unclonable function (PUF), a hardware security primitive for device authentication and cryptographic key generation. Through the presentation, the potential reliability issues and failure mechanisms of these new applications will be discussed.