DUGreen projects for green software development
GreenC5 is a prototype of an adaptive and energy-aware version of the traditional C5 Generic Collection—A well-known comprehensive and open-source data structure library for .NET programming.
The main features of our GreenC5:
- GreenC5 can select and transform to different C5 data structures automatically and dynamically at runtime, for energy efficiency.
- GreenC5 is smart, adaptive, energy-aware and lightweight. The overhead is less than 3%.
- GreenC5 is easy-to-use because the usage is similar to traditional data structures. Programmers do not need to select a data structure. Only one declaration is needed for each instance of the data structures. There is also minimal extra installations and configurations to use GreenC5.
The released projects include:
- Watts Up? Framework: this framework contains APIs for power measurement using a Watts Up? Pro meter.
- Intel Power Gadget Framework: this framework contains APIs for power measurement using the Intel Power Gadget with other required libraries.
- CRUD-Based Collection Energy Profiling: this project contains a program for creating training and program validation datasets using a Watts Up? Pro power meter.
- Machine Learning Framework: this project contains two implementations of an Artificial Neural Network and N-Gram.
- CRUD-Based C5 Collection: this project contains a wrapper/factory class of the C5 data structures.
- Green CRUD-Based C5 Collection: this project contains the GreenC5 data structure class, Green component and other utility classes.
- GreenC5 Simulator: this project is the implementation of our GreenC5 simulator to allow users to interact with the GreenC5 in different use-case scenarios and settings. The project also contains code examples of how the GreenC5 is used; experiment and evaluation projects of the three additional case studies; and the implementations of Watts Up? and Intel Power Gadget based power profilers.
Our publication: http://dx.doi.org/10.1016/j.suscom.2016.11.004
Authors and Contributors: Junya Michanan (@michanaj), Mathematics & Computer Science Department, Chulachomklao Royal Military Academy, Nakhorn Nayok, Thailand
Rinku Dewri & Matthew J. Rutherford, Computer Science Department, University of Denver, Colorado, USA.