Optimizing Probe-Based Storage
FAST '03 Proceedings of the 2nd USENIX Conference on File and Storage Technologies
MEMS-based Storage Devices and Standard Disk Interfaces: A Square Peg in a Round Hole?
FAST '04 Proceedings of the 3rd USENIX Conference on File and Storage Technologies
An overview on MEMS-based storage, its research issues and open problems
SNAPI '04 Proceedings of the international workshop on Storage network architecture and parallel I/Os
Power management of MEMS-based storage devices for mobile systems
CASES '08 Proceedings of the 2008 international conference on Compilers, architectures and synthesis for embedded systems
Optimizing probe-based storage
FAST'03 Proceedings of the 2nd USENIX conference on File and storage technologies
MEMS-based storage devices and standard disk interfaces: a square peg in a round hole?
FAST'04 Proceedings of the 3rd USENIX conference on File and storage technologies
Hi-index | 0.00 |
Power dissipation by storage systems in mobile computersaccounts for a large percentage of the power consumedby the entire system. Reducing the power used by the storagedevice is crucial for reducing overall power consumption.A new class of secondary storage devices based onmicroelectromechanical systems (MEMS) promises to consumean order of magnitude less power with 10-20 timesshorter latency and 10 times greater storage densities. Wedescribe three strategies to reduce power consumption: aggressivespin-down, sequential request merging, and sub-sectoraccesses. We show that aggressive spin-down cansave up to 50% of the total energy consumed by the deviceat the cost of increased response time. Merging of sequentialrequests can save up to 18% of the servicing energy andreduce response time by about 20%. Transferring less datafor small requests such as those for metadata can save 40%of the servicing energy. Finally, we show that by applyingall three power management strategies simultaneously thetotal power consumption of MEMS-based storage devicescan be reduced by about 54% with no impact on I/O performance.