Energy Efficient Video Encoding for Cloud and Edge Computing Instances

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
[email protected] | 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

8. VEEP-Match Architecture
8

9. VEEP-Match Architecture
9

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
MatchVEEP
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.

12. VEEP Submodules
12

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

18. Daily Overview
17
18

19. The Calculator Submodule in VEEP
19

20. VEEP-Match Architecture
MatchVEEP
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.

21. Match Submodules
21

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
[email protected]