Explore the evolution and real-world applications of Oblivious RAM (ORAM) in this 47-minute Google TechTalk presented by Elaine Shi from Carnegie Mellon University. Gain insights into ORAM's journey from a theoretical concept to large-scale deployment, including its emerging demands in blockchain and cloud services. Learn about Signal's successful implementation of Path ORAM, which dramatically reduced their server requirements. Discover the initiative to create an open-source Oblivious STL library and how external-memory algorithm techniques can significantly improve performance for hardware enclaves. Understand the shift in focus from computational overhead to page swaps as a key performance metric in ORAM implementations.