11 July 2026
Montreal, Canada
The
International Telecommunication Union (ITU) and the IEEE 802 LMSC (The IEEE 802 LAN/MAN Standards Committee) organized the
Tenth Joint IEEE 802 and ITU-T Study Group 15 Workshop
– Emerging standards in the AI/ML world which took place on
11 July 2026 in
Montreal, Canada in continuation of this successful series of joint workshops.This 10th edition will be in-person only.This workshop intended to offer a platform for all involved stakeholders and aimed to focus on topics of common interest regarding applications, protocols, higher speed solutions and fiber for artificial intelligence (AI).
The objectives of this workshop included, but are not limited to, enhancing long-standing collaboration and coordination between IEEE 802 and ITU-T Study Group 15 through discussion and information exchange on topics of common interest.
Target Audience
Participation in the workshop was open to ITU Member States, Sector Members, Associates and Academic Institutions and to any individual from a country that is a member of ITU who wishes to contribute to the work. This included individuals who are also members of international, regional and national organizations. Participation in the workshop was free of charge.
🍁 Watch the recording of the session here🍁 Workshop
Photos here
Programme
07:45 - 08:30
|
Registration
|
08:30 - 09:00
|
Opening Remarks
|
09:00 - 10:45
|
Session 1: Introduction – AI Applications
In this session, the ITU-T ION-2030 framework will be introduced. It includes the topics of how AI can be applied to optical networks and how optical networks can be applied to the compute infrastructure supporting AI. From an IEEE 802 perspective, presentations will cover activity on AI Computing Networks within the IEEE 802 Nendica program and its relationship to industry developments. In addition, use cases for wireless LAN, as documented within the IEEE 802.11 Working Group’s activity on AI and Machine Learning, will be presented
|
|
10:45 - 11:15 |
Coffee Break
|
|
11:15 - 12:40 |
Session 2: Protocols
This session explores emerging technologies and standards shaping the next generation of high performance AI datacenter networks.
The first part of this session will address protocol level advances including new mechanisms that enhance reliability and manage congestion enabling massive; and distributed compute fabrics. Topics include technologies used for scale-up, scale-out, and scale-across including Backward Notification (IEEE P802.1EJ), Congestion Signaling (IEEE P802.1EH), technologies standardized by the Ultra Ethernet Consortium, and Metadata Services (IEEE P802.3dt).
The second part of this session will address applications of AI/ML in synchronization networks that enhance timing precision and resilience, it also covers synchronization requirements for AI optimized data centres, and it concludes with a forward looking discussion on standards for intelligent optical transport operations.
-
Eugene Opsasnick, Broadcom - Overview of UEC Link Layer Retry (LLR) and Credit-Based Flow Control (CBFC) [Presentation]
-
Jai Kumar, Broadcom - Scale Across - A Frontier Domain [Presentation]
-
Lihao Chen, Huawei - IEEE P802.1EJ Backward Notification | Don’t tell me what to do, just tell me what’s happening [Presentation]
-
Ramana Reddy Machireddy, Rakuten –
Q13/15, AI/ML in Timing and Sync Networks | AI-based Precision Timing, Resilience, and Network Synchronization [Presentation]
-
Jingfei Lv, Huawei on behalf of
Stefano Ruffini, Calnex Solutions –
Network Synchronization for AI Data Centres | ITU-T Q13/15 and other groups [Presentation]
-
Stephen Shew, Ciena and Liping Chen, CICT- Standards consideration for the intelligent optical transport network operations [Presentation]
|
12:40 - 14:00
|
Lunch Break
|
14:00 - 15:30
|
Session 3: Higher Speed Solutions
The IEEE P802.3dj 200 Gb/s, 400 Gb/s, 800 Gb/s, and 1.6 Tb/s Ethernet project is developing 200 Gb/s electrical and optical PAM4 signaling, as well as 800 Gb/s DP-16QAM for 800 GbE coherent solutions. It is anticipated that AI Backend scale-up, scale-out, and scale-across networks will quickly leverage the technologies being developed in this standard and then begin looking for the next generation of solutions. This panel will review the current state of the IEEE P802.3dj draft standard and explore its impact on the next generation of signaling and potential network solutions.
|
15:30 - 16:00
|
Coffee Break
|
16:00 - 17:30
|
Session 4: Fiber for AI Applications
The rapid expansion of artificial intelligence (AI) workloads is driving unprecedented demands on data center and network infrastructure, exposing fundamental limitations in existing optical fiber technologies. AI training and inference systems require significantly higher interconnect densities, faster data transmission speeds, lower latency, improved energy efficiency, and scalable deployment architectures. At the same time, growing constraints related to data center space, conduit availability, and land utilization are increasing the need for fiber solutions that maximize capacity within a limited physical footprint. Traditional fiber designs, while highly effective for conventional communications networks, are increasingly challenged to support the bandwidth, connectivity, and thermal management requirements of next-generation AI clusters. To address these emerging requirements, new fiber technologies are needed. New fiber technologies are being introduced at Industry conferences, fora and in standards to address these new challenges.
|
