South Korea’s KT has launched dedicated AI network services with bandwidth options up to 800Gbps, expanding from its previous ceiling of about 100Gbps. Seoul Economic Daily reported that the service adds 400Gbps and 800Gbps tiers for large data transfers between distributed AI data centers.

The network uses an APN-based, optics-centered architecture intended to reduce optical-to-electrical conversions along the path. Fewer conversions can reduce latency and simplify long-haul movement of training data, model checkpoints and replicated storage. Yonhap separately reported the launch and the same 100Gbps-to-800Gbps expansion.

Seoul Economic Daily TV recently examined how KT and Korea’s other carriers are expanding AI data-center infrastructure.

KT says its national fiber footprint spans about 840,000 kilometers and can provide shortest-path and redundant routing between sites. The move fits a larger networking shift in which AI infrastructure is demanding much faster links than traditional enterprise traffic. BitcoinVersus.Tech recently covered Cisco’s 51.2 Tb/s Silicon One G302 and Broadcom’s 800G and 102.4 Tb/s OCP networking.

The 800Gbps figure describes line bandwidth, not an automatic application-level transfer rate; actual throughput still depends on endpoints, protocol overhead, congestion and storage performance. The important development is that carrier-grade links are moving closer to the bandwidth class already appearing inside AI clusters, while technologies such as pluggable optics continue spreading into more parts of the network.

BitcoinVersus.Tech Editor’s Note:
We volunteer daily to ensure the credibility of the information on this platform is Verifiably True. If you would like to support to help further secure the integrity of our research initiatives, please donate using the QR code provided

For best QR-code scanning results, please scan this donation code from a desktop or larger screen. Mobile screens may prevent camera recognition.