19.03.2026 - Workshop on Media Energy Consumption Measurement and Exposure
- Mar 12
- 5 min read
Co-organized between 3GPP SA4, Greening of Streaming and 5G-MAG, this workshop will bring experts together to explore the practical integration of energy measurement, reporting, and optimization within media streaming architectures, in particular with a focus on 5G and upcoming 6G systems
REGISTER NOW ! (free attendance)
📅 Date: 19th March from 15:00 to 17:00 CET
📍 Location: On-Line via Zoom
AGENDA
Presentation and Logistics, by Thomas Stockhammer (Qualcomm) and Jordi J. Giménez (5G-MAG)
The 3GPP Work Item " Study on Media Energy Consumption Exposure and Evaluation Framework phase 2" (WI # 1080050), by Julien Lemotheux (Orange) - DOWNLOAD SLIDES
Greening of Streaming's work and perspective, by Benjamin Schwarz (Greening of Streaming) - DOWNLOAD SLIDES
State of energy awareness in today's network management protocols and devices, by Marisol Palmero (Greening of Streaming) - DOWNLOAD SLIDES
Practical Insights from the SMART-CD Project, by Lucas Gregory (ATEME) - DOWNLOAD SLIDES
Practical Insights by Christian Timmerer (Alpen-Adria-Universität Klagenfurt) - DOWNLOAD SLIDES
Ecoflow ECOFLOW: Energy-Conserving Optimization for Future-ready, Low-impact Online Workflows
GitHub resources for reproducibility Reproducibility | ATHENA Christian Doppler (CD) Laboratory
A focus on metrics collection and decision taking
Results from the Green Streaming projects by Martin Lasak (Fraunhofer FOKUS) - DOWNLOAD SLIDES
5G-MAG Reference Tools overview by Daniel Silhavy (Fraunhofer FOKUS, 5G-MAG WG DEV Chair) - DOWNLOAD SLIDES
Open Discussion and Q&A - 30 mins
Moderated by Thomas (Qualcomm) and Jordi (5G-MAG)
Opportunity for participants to share insights, ask questions, and discuss next steps
This is an AI-generated summary of the event, please refer to the slides and let us know if there is anything incorrect by sending an e-mail to: gimenez@5g-mag.com
Summary Per Speaker
Thomas opens the workshop by framing the session as a joint effort between 3GPP, 5G‑MAG, and Greening of Streaming, emphasizing that energy consumption in media delivery has become “an important topic… a design vector for 6G.” He explains that 3GPP is actively studying media‑related energy consumption and is seeking real‑world measurements from industry and research partners. Throughout the session, he manages logistics, transitions between speakers, and encourages questions, aiming to create a collaborative environment for discussing energy‑efficient media delivery.
Julien Lemotheux (Orange / 3GPP SA4 Rapporteur)
Julien provides the operator perspective, explaining Orange’s commitment to achieving net‑zero carbon by 2040 and the challenges that come with it. He notes that while fixed networks have become more efficient—“it was a good thing to switch from ADSL to fiber”—mobile network energy consumption continues to rise despite optimization efforts. A central issue, he argues, is the lack of granular, shareable energy data across devices, networks, and services. Because “we need to share the information… to be efficient on the overall energy reduction,” he introduces the idea of a new 3GPP component, the Energy Information Application Function, designed to bridge network‑side energy data with video players so that energy‑aware decisions can be made end‑to‑end.
Benjamin Schwarz (Greening of Streaming)
Benjamin describes Greening of Streaming’s mission as an evidence‑driven effort to reduce energy consumption in video delivery, stressing that “we only work on things that we can actually measure in watts.” He outlines the organization’s lab structure and its flagship REM (Remote Energy Measurement) project, which measures energy use from encoders through to end‑user devices. He shares several findings: home gateways typically consume around 10 watts continuously regardless of traffic; peer‑to‑peer delivery adds only 0.1–0.2 watts per peer; and enabling eco‑mode on TVs can reduce consumption by roughly 20%. He also highlights that content characteristics—especially brightness—have a far greater impact on device energy use than codec or resolution choices, and that some OLED TVs surprisingly consume less power when playing 4K content than HD.
Marisol Palmero (Greening of Streaming)
Marisol focuses on the standards landscape, arguing that today’s energy‑related metrics are fragmented across encoding, CDN, network, and device domains. She notes that “there is no standard that will give you the measurement… in an end‑to‑end use case,” which limits coordinated energy optimization. She describes ongoing work in the IETF GREEN working group, where a YANG‑based data model is being developed to represent device‑level energy metrics in real time. The model is about 70% complete at the device‑component level, with upcoming work focused on controller interactions and API definitions. She also previews a video‑streaming‑focused hackathon planned for the next IETF meeting in Vienna, intended to advance cross‑SDO collaboration between IETF, 3GPP, and Greening of Streaming.
Lucas Gregory (ATEME)
Lucas presents the SmartCD project, a French collaborative effort led by ATEME to measure and reduce the environmental impact of end‑to‑end streaming workflows. He explains that cloud‑based measurement is particularly challenging because providers rarely expose real energy data, noting that “measuring the environmental impact of streaming workflows in the cloud is really challenging.” To address this, the project relies on resource‑usage metrics—CPU, memory, network traffic, storage activity—combined with energy models. Lucas emphasizes that understanding where energy is consumed is a prerequisite to meaningful optimization and that device generation plays a major role in decoding and rendering energy.
Christian Timmerer (University of Klagenfurt)
Christian shares academic research on energy‑aware encoding, bitrate ladder design, and adaptive bitrate (ABR) algorithms. His work shows that slower encoder presets significantly increase energy consumption with diminishing returns in quality, and that “small quality reductions can still yield substantial energy savings.” He presents energy‑aware bitrate ladders and an energy‑optimized ABR algorithm (eVISH), along with metrics for video complexity, decoding energy, and CO₂ prediction. While industry interest exists, he notes that commercial adoption is limited because customers rarely prioritize or pay for energy‑saving features.
Martin Lasak (Fraunhofer FOKUS)
Martin describes Fraunhofer’s measurement framework, which combines power meters with detailed streaming analytics to understand energy use across devices and workflows. One of the most striking findings is that TV eco‑mode has a dominant impact on energy consumption, yet “many devices were not using echo mode” in real‑world tests. He also confirms that content brightness and device settings outweigh technical parameters like bitrate or resolution in determining energy use. Fraunhofer is exploring LLM‑based user guidance to help viewers reduce energy consumption and is also measuring encoder energy across software and hardware implementations.
Daniel Silhavy (Fraunhofer FOKUS / 5G‑MAG Reference Tools)
Daniel presents the 5G‑MAG reference tools, which implement 3GPP specifications and provide an open‑source environment for experimenting with media delivery, QoE metrics, and client‑side data collection. He explains that “the reference tools can really be a starting point if you want to try something out,” positioning them as a potential foundation for prototyping the Energy Information Application Function proposed earlier in the workshop. The tools already support CMCD, QoE reporting, and UE data collection, making them well‑suited for future energy‑related experimentation.
Relevant References
The 3GPP Study
For insight into the 3GPP SA4 Study: https://www.3gpp.org/ftp/tsg_sa/TSG_SA/TSGS_108_Prague_2025-06/Docs/SP-250636.zip
Outcomes to be captured in TR 26.942: https://www.3gpp.org/dynareport/26942.htm
IETF GREEN Framework Reference Model:
SMART-CD Project - Open-Source Tools:
METRICS & TOOLS
ITU SITI (Spatial Information, Temporal Information): GitHub - VQEG/siti-tools: SI TI calculation tools · GitHub
VCA (Video Complexity Analyzer): VCA | Video Complexity Analyzer
EVCA (Enhanced VCA): EVCA: Enhanced Video Complexity Analyzer | ATHENA Christian Doppler (CD) Laboratory
VEEP: VEEP: Video Encoding Energy and CO2 Emission Prediction | ATHENA Christian Doppler (CD) Laboratory
RDEI: Machine Learning-Based Decoding Energy Modeling for VVC Streaming
Receptive: QoE- and Energy-aware content consumption for HTTP Adaptive Streaming / Daniele Lorenzi
GREEM: An Open-Source Energy Measurement Tool for Video Processing, GREEM | https://doi.org/10.1145/3625468.3652168 | Github
DATASETS
MVCD: Multi-Dimensional Video Compression Dataset: Web | https://doi.org/10.1109/VCIP63160.2024.10849883 | Github
COCONUT: Content Consumption Energy Measurement Dataset for Adaptive Video Streaming: Web | https://doi.org/10.1145/3625468.3652179 | Github
VEED: Video Encoding Energy and CO2 Emissions Dataset for AWS EC2 instances: Web | https://doi.org/10.1145/3625468.3652178 | Github
5G-MAG Reference Tools: https://developer.5g-mag.com/
Greening of Streaming Publications: https://www.greeningofstreaming.org/publications





Comments