Media.Player

Software for Improved Application Performance and Power on Heterogeneous Computers

IP Title

Elastic Computing

Detailed Technology Description

None

Application Date

Jul 23, 2013

Application No.

9,495,139

Others

*Abstract

Invention

This software contains unique optimization algorithms for transparently accelerating processing speeds while using less electricity than alternative solutions. The usage of heterogeneous accelerator devices, such as graphics-processing units and field-programmable gate arrays, has replaced the former computing trend towards faster clock frequencies and additional computing nodes. Although such devices have numerous advantages, their usage and effectiveness remains limited due to significantly increased application design complexity compared to traditional software.

Researchers at the University of Florida have created a software optimization framework, dubbed elastic computing, that draws on implementation alternatives and parallelization strategies for different functions to transparently optimize an application for different types and numbers of computing resources without any designer effort. It works smarter instead of harder, maximizing efficiency for a range of different resources and runtime parameters.

Application

Software that contains optimization and implementation assessment algorithms to make it possible for computers to run calculations faster and at lower power with less designer effort

Advantages

• Transparent optimization algorithms permit adaptation to changing input parameters and resources, delivering better performance and power
• Reduces power and cooling needs compared to traditional computing improvements, saving money and reducing size
• Proven to work on a variety of systems, ensuring maximum versatility
• Hides the complexity of GPU and FPGA application design, enhancing ease of use for non-experts

Technology

This optimization framework automatically and transparently optimizes an application across any type and any number of computing resources. The framework separates functionality from implementation details, permitting designers to use specialized elastic functions that have a similar interface to any function library. These elastic functions enable an optimization framework to explore thousands of possible implementations, even ones employing different algorithms. Additionally, these functions allow efficient execution of the same application code on any architecture according to different runtime parameters, such as input size, available resources, or battery life. This elastic computing framework optimizes application code to achieve significant speedups ranging from 1.3x to 206x, and potentially supports any system architecture.

*IP Issue Date

Nov 15, 2016

*IP Publication Date

Jan 23, 2014

*Principal Investigator

Name: Greg Stitt

Department:

Name: John Wernsing

Department:

Country/Region

USA