Energy Efficient Video Encoding for Cloud and Edge Computing Instances
0 likes122 views
The document outlines the VEEP-Match project, which focuses on energy-efficient video encoding for cloud and edge computing. It aims to optimize video encoding/transcoding operations by balancing cost, energy consumption, CO2 emissions, and encoding time. Results indicate significant savings in both cost and energy across various encoding scenarios.
Energy Efficient Video Encoding for Cloud and Edge Computing Instances
- 1. Energy Efficient Video Encoding for Cloud and Edge Computing Instances Samira Afzal Institute of Information Technology (ITEC), Alpen-Adria-Universität Austria samira.afzal@aau.at | https://athena.itec.aau.at/gaia
- 2. 2 GAIA (the Greek goddess of Earth, mother of all life, personification of the Earth) Funded by the Austrian Research Promotion Agency FFG (Basisprogramm) A cooperative project between Bitmovin and Alpen-Adria-Universität Klagenfurt (AAU) : Intelligent Climate-Friendly Video Platform
- 3. 3 Complete energy awareness and accountability, including energy consumption and GHG emissions along the entire delivery chain, from content creation and server-side encoding to video transmission and client-side rendering Reduced energy consumption and GHG emissions through advanced analytics and optimizations on all phases of the video delivery chain : Intelligent Climate-Friendly Video Platform
- 4. 4 VEEP-Match: Energy Efficient Video Encoding for Cloud and Edge Computing Instances
- 5. Main Objectives To select Cloud/Edge Instances for video encoding/transcoding operations, aiming at: Minimizing cost, energy consumption, CO2 emission or encoding time Making a trade-off between these priorities 5
- 6. Storage/CDN Encoded video Video Coordinator Storage Energy Efficient Video Encoding 6
- 7. Energy Efficient Video Encoding VEEP-Match Storage/CDN Encoded video Video Storage 7 Coordinator
- 10. VEEP VEEP-Match Architecture 10 Lachini, Armin, et al. "VEEP: Video Encoding Energy and CO2 Emission Prediction." Proceedings of the Second International ACM Green Multimedia Systems Workshop. 2024.
- 11. VEEP-Match Architecture Match VEEP 11 Afzal, Samira, et al. "VE-Match: Video Encoding Matching-based Model for Cloud and Edge Computing Instances." Proceedings of the First International Workshop on Green Multimedia Systems. 2023.
- 13. The Video Analyzer Submodule in VEEP 13
- 14. The Predictor Submodule in VEEP 14
- 15. The Calculator Submodule in VEEP 15
- 16. The Data Source Submodule in VEEP 16
- 17. Annual Overview CO₂ emissions over the year 2023 in Austria. 17
- 19. The Calculator Submodule in VEEP 19
- 20. VEEP-Match Architecture Match VEEP 20 Afzal, Samira, et al. "VE-Match: Video Encoding Matching-based Model for Cloud and Edge Computing Instances." Proceedings of the First International Workshop on Green Multimedia Systems. 2023.
- 22. Evaluation 22
- 23. Germany Austria Cheapest with highest energy consumption Most expensive with Lowest energy consumption Experimental Infrastructure 23 500 video encoding segments
- 24. VEEP-Match Experimental Results 77.85% cost reduction 45.42% energy reduction 24
- 25. Conclusions Proposed VEEP-Match for optimized scheduling of adaptive encoding/transcoding processes in cloud and edge environments Proposed a game-theoretic optimization model Highlighted factors: Providers' priorities: cost, energy, CO2 and/or time priorities Resources properties: location, type, cost Encoding parameters: codec, bitrate, resolution Segment features: complexity Evaluated VEEP-Match on a set of cloud AWS and edge resource instances Achieved significant savings: 17%-78% in cost-optimized scenarios 38%-45% in energy-optimized scenarios 25
- 26. Thank you Have a great day ahead! Institute of Information Technology (ITEC) Alpen-Adria-Universität Austria samira.afzal@aau.at