Energy Consumption Analysis of Algorithms Implementations

Rashid, Mohammad Rifat Ahmmad; Ardito, Luca; Torchiano, Marco

doi:10.1109/esem.2015.7321198

Cited by 29 publications

(21 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous studies have claimed that compiled programming languages such as C and C++ are the most energy and performance efficient, while semi-compiled and interpreted languages are the least efficient [1,10,33]. We attempt to investigate whether the above statement is true by conducting a large scale empirical study on a sample of 25 programming tasks implemented in 14 programming languages.…”

Section: Research Questionsmentioning

confidence: 97%

“…Chen and Zong worked on smartphones and showed, by using the Android Run Time environment (instead of Dalvik), that the energy and performance implications of Java are similar to C and C++ [11]. Finally, Rashid et al worked on an embedded system and compared the energy and performance impact of four sorting algorithms written in three different programming languages (arm assembly, C/C++, and Java) [33]. They found that Java consumes most energy and performs slowly against C/C++ and assembly.…”

Section: Across Execution Platformsmentioning

confidence: 99%

See 1 more Smart Citation

What are your programming language's energy-delay implications?

Georgiou

Kechagia

Λουρίδας

et al. 2018

Proceedings of the 15th International Conference on Mining Software Repositories

View full text Add to dashboard Cite

Motivation: Even though many studies examine the energy efficiency of hardware and embedded systems, those that investigate the energy consumption of software applications are still limited, and mostly focused on mobile applications. As modern applications become even more complex and heterogeneous a need arises for methods that can accurately assess their energy consumption. Goal: Measure the energy consumption and run-time performance of commonly used programming tasks implemented in different programming languages and executed on a variety of platforms to help developers to choose appropriate implementation platforms. Method: Obtain measurements to calculate the Energy Delay Product, a weighted function that takes into account a task's energy consumption and run-time performance. We perform our tests by calculating the Energy Delay Product of 25 programming tasks, found in the Rosetta Code Repository, which are implemented in 14 programming languages and run on three different computer platforms, a server, a laptop, and an embedded system. Results: Compiled programming languages are outperforming the interpreted ones for most, but not for all tasks. C, C#, and JavaScript are on average the best performing compiled, semi-compiled, and interpreted programming languages for the Energy Delay Product, and Rust appears to be well-placed for i/o-intensive operations, such as file handling. We also find that a good behaviour, energywise, can be the result of clever optimizations and design choices in seemingly unexpected programming languages. CCS CONCEPTS • Hardware → Power estimation and optimization; • Software and its engineering → Software libraries and repositories; Software design tradeoffs;

show abstract

Section: Research Questionsmentioning

confidence: 97%

Section: Across Execution Platformsmentioning

confidence: 99%

What are your programming language's energy-delay implications?

Georgiou

Kechagia

Λουρίδας

et al. 2018

Proceedings of the 15th International Conference on Mining Software Repositories

View full text Add to dashboard Cite

show abstract

“…One example for this latter type of analysis is the energy efficiency assessment of different programming languages and algorithms to solve a specific problem, e.g. the work of presented by Rashid [34] considering sorting algorithms. In a study by Johann et al, source code instrumentation is used to locate resource intensive parts of programs in order to improve them [20].…”

Section: Related Workmentioning

confidence: 99%

Code-Level Energy Hotspot Localization via Naive Spectrum Based Testing

et al. 2018

View full text Add to dashboard Cite

With the growing adoption of ICT solutions, developing energy efficient software becomes increasingly important. Current methods aimed at analyzing energy demanding portions of code, referred to as energy hotspots, often require adhoc analyses that constitute an additional process in the development life cycle. This leads to the scarce adoption of such methods in practice, leaving an open gap between source code energy optimization research and its concrete application. Thus, our underlying goal is to provide developers with a technique that enables them to efficiently gather source code energy consumption information without requiring excessive time overhead and resources. In this research we present a naive spectrumbased fault localization technique aimed to efficiently locate energy hotspots. More specifically, our research aims to understand the viability of spectrum based energy hotspot localization and the tradeoffs which can be made between performance and precision for such techniques. Our naive yet effective approach takes as input an application and its test suite, and utilizes a simple algorithm to localize portions of code which are potentially energy-greedy. This is achieved by combining test case coverage information with runtime energy consumption measurements. The viability of the approach is assessed through an empirical experiment. We conclude that the naive spectrum based energy hotspot localization approach can effectively support developers by efficiently providing insights of the energy consumption of software at source code level. Since we use processes already in place in most companies and adopt straightforward data analysis processes, naive spectrum based energy hotspot localization can reduce the effort and time required for assessing energy consumption of software and thus make including the energy consumption in the development process viable. As future work we plan to (i) further investigate the tradeoffs between performance and precision of spectrum based energy hotspot approaches (ii) compare our approach to similar ones through large-scale experiments. Our ultimate goal is to conceive ad-hoc tradeoff tuning of performance and precision according to development and organizational needs.

show abstract

“…Several pieces of research have measured power consumption directly using instrumentation hardware. This can be in terms of the overall system power [26], which can be measured by the current drawn from the battery on mobile devices, such as in [19] and in the "Green Miner" platform [20]. Power consumption due to the CPU can also be determined via the Intel Power Gadget API [2], [27].…”

Section: Related Workmentioning

confidence: 99%