Cutting the Cost and Power Consumption of Big Data Applications, in a Flash

It’s possible to cut the cost and power consumption of Big Data applications by using flash memory as a replacement for conventional RAM, according to new research from ISTC Principal Investigator Arvind and fellow researchers from MIT CSAIL, MIT, and Quanta Research Cambridge, as reported by the MIT News Office on July 10.  In a paper presented at the recent International Symposium on Computer Architecture (ISCA 2015), the researchers described a new system architecture that, for several common big-data applications, should make servers using flash memory (a common form of memory used in portable devices) as efficient as those using conventional RAM, while preserving their power and cost savings.

According to the researchers, the new system architecture, called BlueDBM, has flash-based storage with in-store processing capability and a low-latency high-throughout inter-controller network. “We show that BlueDBM outperforms a flash-based system without these features by a factor of 10 for some important applications,” the researchers wrote.  “BlueDBM presents an attractive point in the cost-performance trade-off for Big Data applications.”

Arvind and students/lead paper authors Sang-Woo Jun and Ming Liu provided a preview of the technology behind BlueDBM in a blog post here in November 2013.

The architecture of BlueDBM, a fast distributed storage system based on flash storage for speeding Big Data analytics.  (Courtesy of Arvind, MIT CSAIL.)

The architecture of BlueDBM, a fast distributed storage system based on flash storage for speeding Big Data analytics. (Courtesy of Arvind, MIT CSAIL November 2013

Resources:

 

 

This entry was posted in Analytics, Big Data Architecture, Computer Architecture, ISTC for Big Data Blog and tagged , , , , , , . Bookmark the permalink.

Leave A Reply

Your email address will not be published. Required fields are marked *


4 + = eleven