Phase Change Memory-based Hardware Accelerators for Deep Neural Networks
- Geoffrey Burr
- Pritish Narayanan
- et al.
- 2023
- VLSI Technology 2023
I am leading the delivery of our on-prem AI systems which will bring our Vela architecture and the Artificial Intelligence Unit to Research partners. These systems are designed to provide an optimized full-stack solution for generative AI and efficiently serve the full AI lifecycle from data pre-processing and training to tuning and inferencing/serving. Prior to this role I was the Chief of Staff and Technical Assistant to Mukesh Khare, GM IBM Semiconductors and VP Hybrid Cloud Research, and helped to shape the business and technical strategy for advanced infrastructure research at IBM. I am a semiconductor technologist and have driven cross-functional global teams to deliver advanced prototype hardware for analog in-memory computing as part of IBM’s AI Hardware Center. The output of this work has recently been published in Nature and Nature Electronics. I joined IBM in 2011 as a member of the Advanced Patterning team and became an expert in lithography solutions including multi-patterning lithography, EUV lithography, OPC, and source mask optimization for 14nm technology and beyond. Throughout my career at IBM, I have held numerous management and leadership roles and also co-developed an internal program for skill growth. I earned my Ph.D. in Electrical and Computer Engineering from Carnegie Mellon University where I studied computational modeling of single photon emitting devices.
Additional research interests include: analog in-memory computing, semiconductor process development, and generative AI for enterprise