Webinar: What's next in on-device generative AI
The generative AI era has begun. Upcoming trends in generative AI and transformers, from large multimodal models to transformers extending to more modalities, are set to revolutionize user experiences, disrupt industries, and offer new capabilities. For generative AI to reach its full potential, bringing these capabilities to edge devices where they run locally is essential for improved latency, pervasive interaction, and enhanced privacy. This presentation will explore how Qualcomm's end-to-end system philosophy is at the forefront of enabling this next wave of innovation on the edge, including:
Efficiency techniques to bring LLMs, LVMs, and LMMs on device
Utilization of heterogeneous computing
Adaptation and personalization techniques
LLM orchestration at the edge
About the speaker:
Dr. Joseph Soriaga
Senior Director of Technology
Dr. Joseph Soriaga is a Senior Director of Technology and leads the core team within Qualcomm AI Research that is broadly responsible for driving hardware and software platform innovations, advancing applications in areas such as mobile, automotive, XR and wireless, and contributing to fundamental research in machine learning and generative AI. Prior to this role, he led 5G core physical layer research team within Qualcomm Wireless R&D and helped drive the standardization of 5G as a both a 3GPP RAN1 delegate and a technical specifications editor. He has also had previous experience as systems engineering lead for Qualcomm’s 3G EV-DO commercial cell-site modem (CSM). He received his BSEE from Caltech, and PhD ECE from UCSD with focus on coding theory, communications, and information theory. He was awarded IEEE Data Storage Best Paper in 2007 and the EURASIP Journal on Wireless Communications and Networking Best Paper in 2012, and he has been granted over 280 US Patents.