Paper Title: Introduction to Big Memory

Paper Abstract: Persistent memory is here today. It gives system designers storage at memory speeds rather than through relatively slow interfaces. Sounds great, but how do you manage it? Big memory management allows you to handle persistent memory in the same familiar way as storage with similar structures, commands, and tools. Now you can have storage systems with multiple tiers ranging from hard drives through ultra-fast DRAM with persistent memory being much faster than drives but more expensive and much slower than DRAM but cheaper. Initial software implementations are available for evaluation. The results should be greatly improved cost-performance tradeoffs for such applications as databases, financial transactions, AI training, and VR/AR.

Paper Author: Chuck Sobey, Chief Scientist, ChannelScience

Author Bio: Chuck Sobey is an internationally-respected technology advisor, researcher, and lecturer, as well as the General Chairperson of the Flash Memory Summit. He is the founder of the R&D services firm ChannelScience, which develops new capabilities in data storage. Currently, he is advising on such major trends as STT-MRAM, ReRAM, memory manufacturing in China, computational storage, 5G, AI/ML acceleration, and standards for personal AI agents. He has also provided strategic storage technology advice to government agencies worldwide. He has presented at many conferences, including Flash Memory Summit, and has given courses on storage-related topics around the world. He earned electrical and computer engineering degrees from Carnegie Mellon University and the University of California at Santa Barbara.