Parametrize
From Green software
(→Sources) |
|||
Line 13: | Line 13: | ||
== Sources == | == Sources == | ||
- | This best practice was recognized as such by IT professionals, described in [[Energy efficient software (Master Thesis)|''Energy efficient software'']]. | + | This best practice was recognized as such by IT professionals, described in [[Energy efficient software (Master Thesis)|''Energy efficient software'']].<br /> |
+ | This best practice is mentioned in:<br /> | ||
+ | Baek, W. & Chilimbi, T. M. 2010. Green: a framework for supporting energy-conscious programming using controlled approximation. SIGPLAN Not., ACM, 45, p. 198-209 |
Current revision
Quality of service (QoS) is a set of properties that reflect the user experience. These properties include amongst others availability, reliability and performance. To achieve a high quality of service usually more energy has to be consumed as opposed to a low quality of service.
By reducing the availability of a system from 99.999% to 99.9% can mean there is need for less redundant servers which saves a lot of energy. On other quality properties energy can be saved as well by reducing them, like reliability and performance.
Implementations
The Green framework, developed at Microsoft, has been developed to reduce QoS while guaranteeing a certain level of QoS. It is an example of how the execution of an algorithm can be cut short to increase the energy efficiency at the cost of quality. This framework allows any programmer to make the trade-off happen at runtime.
References
The paper on the Green Framework: http://research.microsoft.com/en-us/um/people/trishulc/papers/green.pdf
Energy Consumption Reduction with Low Computational Needs in Multicore Systems with Energy-Performance Tradeoff http://www.nt.ntnu.no/users/skoge/prost/proceedings/cdc09/data/papers/2817.pdf
Sources
This best practice was recognized as such by IT professionals, described in Energy efficient software.
This best practice is mentioned in:
Baek, W. & Chilimbi, T. M. 2010. Green: a framework for supporting energy-conscious programming using controlled approximation. SIGPLAN Not., ACM, 45, p. 198-209