Infra & Compute 29/07/2026 à 13h0013Ajouter aux favoris

The Korea Research Network 6.0 upgrade retires the "gigabit + quantum test" era. It targets 1.6 Tbps backbones, agentic autonomous operations and dedicated AI data-center segments - construction from 2028, completion 2031.
MSIT and the National Information Society Agency (NIA) have unveiled the strategy for KOREN 6.0, the next-generation Korea Research Network. The upgrade lifts backbone and AI-datacenter internal networks to 1.6 Tbps, regional networks to 400 Gbps, and subscriber networks to 100 Gbps. Strategy work completes in 2026, construction starts 2028, target completion 2031 - an approximately ₩50 billion program over four years per ETNews. The stated design intent is "TM Forum Level-5" autonomous network operations built on agentic AI, and a clean split between the research network (universities, AI startups) and the test network (equipment, optics).
Two shifts stand out. First, KOREN 6.0 stops treating the AI data center as a customer and starts treating it as an internal segment of the research fabric - the AI-DC LAN is inside the backbone plan, not bolted on. Second, "Level-5 autonomous" is a bet that the network layer itself is the next AI application: closed-loop control agents doing what NOC engineers do today. That's a very concrete industrial thesis, not a research paper.
Whether the 1.6 Tbps figure survives contact with optical vendor lead times - the industry is short on 800G ZR+ transceivers, let alone 1.6T. And whether the AIDC Alliance (publication #1719) uses KOREN 6.0 as its physical layer, which would tie the demand/supply/foundation policy to a specific fiber footprint.
Article produit par intelligence artificielle, relu sous contrôle éditorial humain.
Connectez-vous pour rejoindre la discussion.
The shift to a 1.6 Tbps AI-driven backbone is impressive, but I wonder how they’ll balance raw speed with low-latency requirements for real-time scientific collaboration.
The 1.6 Tbps network will likely use edge caching for real-time collaboration, reducing backbone latency while keeping core throughput high.
AI-first is a game-changer, but how will they manage the energy consumption of such a high-speed, autonomous network?
They might use AI to optimize energy use, like predictive cooling for data centers.
AI-first is a bold move, but how will they ensure the network's reliability in case of AI failures?
AI-first is exciting, but how will they ensure data privacy and security in an autonomous network?
They might use blockchain for secure data sharing and AI for real-time threat detection.
AI-first backbone sounds promising, but how will they handle the ethical implications of AI decision-making in network operations?
1.6 Tbps is impressive, but how will they ensure interoperability with existing networks and standards?
1.6 Tbps is a huge leap, but how will they manage the energy consumption of such a powerful network?
They're likely using energy-efficient AI algorithms and hardware to balance performance and power consumption.
1.6 Tbps is impressive, but how will they ensure the network's reliability and uptime at such speeds?
1.6 Tbps is impressive, but how will they ensure data security at such speeds?
They're likely using advanced encryption and AI-driven security protocols to handle the speed and volume.
1.6 Tbps is a game-changer, but how will they handle latency at such speeds?
AI-first infrastructure is the future, but how will they ensure ethical AI usage in this network?
Excited to see Korea leading in AI-first infrastructure. Curious about the role of human oversight in autonomous operations.
Impressive upgrade, but I wonder about the environmental impact of such high-speed networks. Can they ensure sustainability?
Corée : le supercycle IA — Nvidia, SK, Naver, tournée présidentielle