Improving Operating Systems by Replacing Hard Disks with Persistent Solid State Memory

University Of California-Los Angeles, Los Angeles CA

Investigators

Abstract

The use of persistent RAM in consumer products has increased its capacity and decreased its price. These trends suggest that typical computers could contain large quantities of this kind of memory in the near future. This project will investigate how to change operating system design for workstation and server machines to make best use of the availability of large quantities of persistent RAM. In particular, the project will investigate the use of persistent RAM for a machine's primary stable storage, replacing hard disks. The project will perform a test design and implementation to demonstrate the feasibility of the concept and to suggest possible advantages of the approach. This implementation will be performed on a Linux system augmented with 2 Gbytes or more of RAM that will serve as its primary storage device, with a hard disk used only for large files, such as video and audio data. The project will demonstrate advantages in speed and simplicity of the system. It will also point out interesting possibilities for improvements in operating system services made possible by the use of persistent RAM.