ION-2030 for AI Data Centers: From Scale-Up to Scale-Out, Scale-Across, and Distributed Intelligence

Tuesday, 3 November 2026 | 8:30 a.m.–12:00 p.m.

 

Session Description:

The International Telecommunication Union (ITU) has established a new strategic framework, International Optical Networks Towards 2030 and Beyond (ION-2030), to guide the evolution of global optical connectivity in the era of artificial intelligence (AI). ION-2030 presents a strategic vision for how optical networks should evolve to meet the emerging requirements of AI, data centers, broadband access, home networking, and IMT-2030/6G mobile networks. The framework emphasizes the mutual empowerment of AI and optical networking, positioning them as complementary drivers of the next wave of digital transformation and ubiquitous intelligence.


This Special Symposium will bring together experts at ACP 2026 to present and discuss advances in ION-2030 technologies and architectures for AI data centers, with particular emphasis on the following four dimensions:

·  Scale-up: Ultra-high-bandwidth and ultra-low-latency connectivity among accelerators, memory, storage, and network interfaces within AI supernodes;

·  Scale-out: Scalable interconnection among racks, pods, and accelerator clusters within a data center or campus;

·  Scale-across: Deterministic, high-capacity connectivity among geographically distributed AI data centers operating as a unified computing infrastructure; and

·  Distributed intelligence: Coordinated AI inference, training, data processing, and resource orchestration across centralized, regional, edge, enterprise, and user-premises computing facilities with a supporting optical network fabric.


Optical technologies will play a critical role across all four dimensions, and the ION-2030 Special Symposium will address key topics including:

·  Co-packaged optics (CPO), near-packaged optics (NPO), eXtra-dense pluggable optics (XPO), linear pluggable optics (LPO), and high-density interconnects;

·  Optical circuit switching (OCS) for AI supernodes, scale-out fabrics, and multi-data-center connectivity;

·  400G, 800G, 1.6T, Beyond-1-Terabit OTN, flexible-grid, full-spectrum transponder (FST), and multi-rail transmission;

·  Deterministic latency, low jitter, lossless or near-lossless transport, and telecom-grade protection and restoration;

·  Distributed training, collaborative inference, agentic AI, embodied AI, and edge intelligence;

·  AI-ready data-center access, metro computing networks, and one-hop connectivity to edge and regional AI data centers;

·  Energy-efficient architectures, sustainability, resilience, security, data sovereignty, and assured service-level agreements; and

·  Standards, deployment experience, technology roadmaps, and emerging business opportunities.


The symposium will bring together experts from network and cloud operators, AI data-center providers, equipment and component manufacturers, standards organizations, universities, and research institutes. Through invited presentations and forward-looking discussions, participants will examine how advances across devices, systems, networks, computing architectures, and intelligent control can be brought together to build scalable, sustainable, secure and resilient AI infrastructure.

We warmly invite the ACP community and professionals across the AI, data-center, and optical-networking ecosystems to join this timely discussion and help shape the evolution of ION-2030 for AI data centers - from scale-up and scale-out to scale-across and truly distributed intelligence.


Moderators:

Shikui Shen, China Unicom, China

Xiang Liu, Hong Kong Research Center of Huawei, Hong Kong, China

 

Agenda:

          Session 1: 8:30 a.m.–10:00 a.m.

1. Glenn Parsons, Ericsson, Canada (ITU-T SG15 Chair)

ITU-T SG15’s Vision and Work Programs for the AI Era

2. Raul Muñoz, Centre Tecnològic de Telecomunicacions de Catalunya (CTTC), Spain

ION-2030 for Distributed AI Training and Inference

3. Nicola Calabretta, Eindhoven University of Technology, Netherlands

Optical Circuit Switching (OCS) for Energy-Efficient AI Data Centers

4..Yuyang Liu, China Telecom, China

Metro Optical Networks Rearchitected as Computing Power Interconnect for Edge-AI

5. Mingqing Zuo, China Mobile, China

An Operator's Perspective on the Evolution towards AI-OTN

6. Gabriele di Rosa, Adtran Networks SE, Germany

Optical Technologies for 1.6Tb/s Coherent Pluggable Modules

-Panel Discussion and Q&A of Session 1

Coffee break: 10:00 a.m.–10:30 a.m.

Session 2: 10:30 a.m.–12:00 p.m.

7. Frank Chang, Source Photonics, USA

1.6Tb/s Optical Modules for AI Data Center Scale-Out

8. Son T. Le, Ciena Corporation, United States

Optical Technologies for AI Data Center Scale-Up

9. Jerry Chen, Alibaba, China

CPO and NPO for Hyper-Scale AI Data Centers

10. Sidong Fu, Tencent, China

NPO and XPO for AI Data Centers

11. Sisi Tan, Huawei, China

Full-Spectrum Transponder Innovations for AI Data Center Scale-Across

-Panel Discussion and Q&A of Session 2


Presented By:

ITU.png

HOSTS
WESTLAKE UNIVERSITY

WESTLAKE UNIVERSITY

ZHEJIANG UNIVERSITY

ZHEJIANG UNIVERSITY

TECHNICAL SPONSORS
IEEE Photonics Soceity

IEEE Photonics Soceity

OPTICA

OPTICA

SPIE

SPIE

THE CHINESE OPTICAL SOCIETY

THE CHINESE OPTICAL SOCIETY

CHINA INSTITUTE OF COMMUNICATIONS

CHINA INSTITUTE OF COMMUNICATIONS

CHINA INSTITUTE OF ELECTRONICS

CHINA INSTITUTE OF ELECTRONICS

POC

POC

CONFERENCE SUPPORT
Learning Conference

Learning Conference

BRIGHTEN CORE

BRIGHTEN CORE