Streamlining Big Data Access

March 7, 2017

Headshot of Nihata AliparmakDr. Nihat Altiparmak, assistant professor of computer engineering and computer science, was awarded a CISE Research Initiation Initiative (CRII) research grant of $175,000 from the National Science Foundation (NSF).

The grant will support his research on self-optimizing high performance data storage systems. The research project looks to develop novel, theoretically grounded, and experimentally validated methods for online detection and automatic elimination of storage performance bottlenecks of data intensive applications.

Optimizing the input and output

Simply put, Dr. Altiparmak’s research is meant to simplify the often-unwieldy process of cloud data recovery. As it stands, files are spread out across various hard drives found on server farms, often in disparate locations from one another, and at a variety of locations internal to each hard drive.

From complex research data to streaming media platforms, By analyzing data usage patterns, Dr. Altiparmak’s optimization process will codify the flow of complex data, a quicker and more efficient system for aiding in climate simulations, drug discover, numerical computations, space discovery, to name a few.

“In real-time while these requests are being made by users, we are analyzing these requests and trying to come up with patterns that we can use for optimization uses,” said Dr. Altiparmak. “The thing that is missing right now is what optimization means. What we are interested in is providing a faster service to users. By analyzing these patterns, we want the users to experience faster. It’s like piggybacking and understanding those requests and analyzing them and coming up with an optimization, and applying them.”

He adds, “Storage systems are shared by multiple users and multiple applications by these users. So eventually there are multiple requests going to these users. You can see two applications interacting with each other. The requirement for our application is generally data driven. These are very costly applications. The computation wise, it takes a very long time to run this simulation (economics, weather patterns). If we can eliminate that bottle neck, we can optimize.”

Receiving the CISE Award

The receipt of a CISE, or Computer and Information Science and Engineering is an accomplishment for anyone and one that Dr. Altiparmak has worked diligently to obtain.

According to the NSF website, “to achieve these, CISE supports investigator initiated research in all areas of computer and information science and engineering, fosters broad interdisciplinary collaboration, helps develop and maintain cutting-edge national computing and information infrastructure for research and education, and contributes to the development of a computer and information technology workforce with skills essential for success in the increasingly competitive global market.”

Dr. Altiparmak explains of his efforts, “I think there are 8 or 9 directorates in NSF. Like biological sciences, engineering, education. It is generally, you are of course competing with top universities. Acceptance rates are reported at 20%. Your competitors or rivals, are general big shots. Everyone is applying there. NSF is the most prestigious place to get an award from. It’s pretty tough to get.”

Helping the Environment

Relative to his optimization algorithms, is the improved efficiency of the various server farms that utilized to store cloud based information throughout the world. The amount of energy necessary to maintain each center is remarkable. As such, his research will not only expedite computing, saving manpower hours in data recovery, but have the secondary benefit of energy conservation. According to Dr. Altiparmak, “Just US data centers are equivalent to all of the energy usage in Kentucky. This is the cost in terms of energy.”