Future non-volatile memory systems

 

One of the challenges in enterprise storage and server systems is the rapidly widening gap between the performance of the hard-disk drives (HDD) and that of the remainder of the system. Research and development efforts have recently shifted towards semiconductor memory technologies that not only complement the existing memory and storage hierarchy but also reduce the distinction between memory (fast, expensive, volatile) and storage (slow, inexpensive, nonvolatile).

The current cost per GB and latency characteristics of NAND Flash make it an interesting candidate for enterprise storage systems. In particular, the introduction of multi-level cells (MLC) further decreased the cost compared to single-level cells (SLC), albeit at the expense of reliability and performance degradation, which is often not acceptable for enterprise applications.

Part of our work focuses on the reliability and performance enhancement of NAND Flash memory in order to enable its employment in enterprise systems. Although NAND Flash has already established itself in the memory and storage hierarchy between DRAM and HDDs, it still cannot serve as a universal memory/storage technology.

There are various other solid-state technologies that combine the performance benefits of semiconductor memories with the low cost of magnetic storage. These include ferroelectric, magnetic, phase-change, and resistive random-access memories. Of those, phase-change memory (PCM) is in the most advanced state of development today.

Our research group is working on various aspects of PCM technology, focusing on the enablement of MLC storage, to make the technology more cost-competitive.