This lecture covers the landscape of big data, the importance of memory in online services, and the challenges faced by memory systems. It discusses the wishlist for a memory system, the key challenge of bandwidth, and the memory hierarchy. The lecture explores emerging DRAM technologies, die-stacked caches, and page-based DRAM caches. It also delves into the implications of skewed dataset accesses, footprint prediction, and the performance comparison of different DRAM caches. The lecture concludes with discussions on storage-class memory, the integration of DRAM and SCM, and the design guidelines for SCM-DRAM systems.