When bridging the computational gap between edge devices and deep learning algorithms, it is necessary to optimize across the hardware, model architecture and software layers, in order to achieve reliable, low latency and energy-efficient performance. For vision specifically, where accuracy is usually very important, trade-offs between model performance and energy efficiency can dictate many engineering choices. This panel will look into how accelerators, if properly integrated, can solve some of these problems, and cover progress at the model architecture level (i.e. compression, pruning, multimodal inferencing) that is helping to deliver high performance, low latency, and energy-efficient inference at the edge.
Jeff White
Jeff is the Industry CTO of Dell Technologies of the Automotive sector, specifically in the area of Connected and Autonomous Vehicles and overall Edge Technology strategy lead. Jeff is responsible for leading the team that is developing the overall Dell Technologies technology strategy development, architectural direction and product requirements for the Intelligent Connected Vehicle platform.
He also is the Chairman of the Dell Automotive Design Authority Council responsible for the technical solution design. In his role as Edge Technology Lead he is driving the development of a Dell Technology wide Edge platform including the physical edge systems, heterogenous compute, memory/storage, environment, security, data management, control plane stack and automation/orchestration.
Previously, Jeff has held senior roles at an early stage artificial intelligence/machine reasoning based process automation technology provider and Elefante Group a stratospheric wireless communications platform as CTO. He also held senior positions at Hewlett Packard Enterprise, Ericsson and Alcatel-Lucent where he led technology initiatives, solutions development, business development and services delivery.
Earlier in his career White served in leadership roles at BellSouth and Cingular Wireless (now AT&T). At Cingular, he led the National Transport Infrastructure Engineering with responsibility for national transport, VoIP & IMS engineering. At BellSouth (now AT&T) he led the Broadband Internet Operations & Support organization which included broadband access tier two technical support, customer networking equipment business, broadband OSS & end-to-end process.
White holds a Bachelor of Science degree in Electrical Engineering from Southern Polytechnic University. He also served as Chairman of the Tech Titans Technology Association of North Texas representing over 300 Technology companies in the greater North Texas community. He also served on the North Texas Regional committee of the Texas Emerging Technology fund under Governor Rick Perry.
Keith Kressin
I currently serve as CEO of MemryX, Inc., an innovative Edge AI startup founded by professors at the University of Michigan in Ann Arbor. Dr Wei Lu is our co-founder and current CTO, and an expert in memories and neuromorphic computing. We plan to enter production in 2023 with the world's best Edge AI solution.
Prior to MemryX, I worked at Qualcomm for 13+ wonderful years, leading a variety of activities including the Snapdragon roadmap, all technologies for application processing, company strategy, competitive analysis, and more. I joined just prior to the first Android smartphone being launched. My most recent role was as SVP/GM helping Qualcomm diversify from smartphones, and leading several growth businesses including AR/VR, AI, PCs, and gaming. I very much enjoyed working with US partners like Facebook, Microsoft and Google in addition to international customers like Vivo, Oppo, Xiaomi, Sony, Samsung, and so many more.
Prior to Qualcomm, I was able to work at Intel for 8+ years in a number of technology, product, and planning roles in different market segments including laptops, desktops and servers. Together we accomplished a # of industry firsts.
During my time at Intel and Qualcomm, I was deeply engaged in shipping substantially more than 1 Billion advanced chips for mobile personal computing devices. I look forward to helping to grow the use of Edge AI in a number of clients as the world continuously becomes both more intelligent and more connected.
Beyond the world of technology, working with smart/dedicated/honest/ethical people is what keeps me enthused to work each day.
Vinesh Sukumar
Vinesh Sukumar currently serves as Senior Director – Head of AI/ML product management at Qualcomm Technologies, Inc (QTI). In this role, he leads AI product definition, strategy and solution deployment across multiple business units.
•He has about 20 years of industry experience spread across research, engineering and application deployment. He currently holds a doctorate degree specializing in imaging and vision systems while also completing a business degree focused on strategy and marketing. He is a regular speaker in many AI industry forums and has authored several journal papers and two technical books.
Pushpak Pujari
Pushpak leads Product Management at Verkada where he runs their Cloud Connected Security Camera product lines. He is responsible for using AI and Computer Vision on the camera to improve video and analytics capabilities and reduce incidence response time by surfacing only meaningful events in real-time, with minimal impact on bandwidth.
Before Verkada, Pushpak led Product Management at Amazon where he built the end-to-end privacy-preserving ML platform at Amazon Alexa, and launched a no-code platform to design and deploy IoT automation workflows on edge devices at Amazon Web Services (AWS). Previous to Amazon, he spent 4 years at Sony in Japan building Sony’s flagship mirrorless cameras.
Pushpak has extensive experience of starting, running and growing multi-million dollar products used by millions of users at the fastest growing companies in the US and the world. He holds an MBA from Wharton and Bachelors in Electrical Engineering from IIT Delhi, India
Sandeep Garg
Sandeep Garg is the Director of Engineering for Catapult at Siemens EDA, responsible for our High-Level Synthesis (HLS) and verification products.
Sandeep has more than 26 years of experience in EDA industry, 18+ of which have been HLS and verification at Siemens, Mentor Graphics and Calypto in various leadership roles in engineering. Catapult’s bottom-up and low-power solutions were launched under Sandeep’s leadership. Sandeep has been instrumental in building up an extensive network of partnerships and ecosystem with academia and research institutes around Catapult HLS, helping evangelize the HLS tools and methodologies and collaborating with cutting edge research initiatives. Before HLS, Sandeep was an architect and a technology and program manager for Mentor Graphics’ Precision FPGA synthesis product line, helping replace their old HDL frontend with a modern alternative and contributing significantly to both its frontend and backend optimizations.
Sandeep came to Mentor Graphics via IKOS Systems acquisition where he was working on hardware-assisted co-simulation products and RTL compiler technology that have since been adopted by many SEDA products.