Hao Zhong & Avi Shetty & David Flynn | Robotics & AI Infrastructure Leaders
AI Summary
This video features a panel discussion at theCube Robotics and AI Leaders series, focusing on the evolving data platform infrastructure supporting AI, robotics, and autonomous technologies. Panelists include Avi Shetty from Solidigm, David Flynn from Hammerspace, and Hao Zhong from ScaleFlux. Key topics include the transformation of storage from traditional hard drives to flash-based capacity tiers, the critical role of high-density, low-latency, scalable SSDs optimized for AI data centers, and innovations like liquid-cooled SSDs. They discuss how AI-native storage architectures aim to reduce data transmission hops to improve performance and lower power consumption, emphasizing the trade-offs between compute and storage tiers. The experts foresee a significant shift in data center architecture, including the disintermediation of storage servers, the importance of network-attached flash, and the increased need for efficient data pipelines to support large-scale AI workloads and robotics. The panel also touches on the future impact of robotics and AI agents on productivity and speculates on the timeframe for robots surpassing humans in population, with predictions around five years. The discussion highlights ongoing breakthroughs in storage density, performance, power efficiency, and system architecture shaping the future AI infrastructure landscape.