Basics of Nonvolatile Memories: MRAM, RRAM, and PRAM - Presented by Fatih Hamzaoglu

Abstract: NAND Flash and eFlash have been the workhorse of memory hierarchy for Standalone Storage and Embedded Non-Volatile Memories, respectively. But with the ever increasing need for memory capacity and bandwidth, due to new applications in Graphics, AI and IoT, device and circuit designers have been heavily investigating alternative memories to fill the need. On the Storage/Memory side, there’s a big gap between DRAM and NAND in terms of density and speed that would justify a new memory device. On the embedded NVM, growing IoT requires better performance and power efficient eNVM than eFlash which can be scaled to 1xnm technologies. This tutorial will talk about basic characteristics, circuit and systems design for emerging non volatile memories such as MRAM, RRAM, and PRAM. The tutorial will go through silicon learning for MRAM and RRAM as well as comparison of these memory devices, and which applications they target. Bio: Fatih joined Logic Tech. Development at Intel after finishing the Ph.D. in EE at the University of Virginia in 2002. Since then, he has been working on memory developments, such as SRAM, DRAM, MRAM and RRAM. He has served as technical committee member at VLSI Symp. Circuits and ISSCC. He has co-authored more than 40 papers and he's co-inventor of more than 30 patents.

• #education
• #mram
• #rram
• #pram
• #nonvolatile_memories