{"id":48561,"date":"2026-06-10T18:29:08","date_gmt":"2026-06-10T18:29:08","guid":{"rendered":"https:\/\/www.europesays.com\/korea\/48561\/"},"modified":"2026-06-10T18:29:08","modified_gmt":"2026-06-10T18:29:08","slug":"ntt-sk-group-bet-500m-on-photonics-to-solve-ai-power-drain","status":"publish","type":"post","link":"https:\/\/www.europesays.com\/korea\/48561\/","title":{"rendered":"NTT, SK Group bet $500M on photonics to solve AI power drain"},"content":{"rendered":"

NTT<\/a>, SK Group<\/a>, and Chunghwa Telecom<\/a> have pooled $500 million into a fund for AI infrastructure built on the IOWN photonics architecture.<\/p>\n

The vehicle \u2013 IOWN AI Fund, managed by the newly-incorporated IOWN Global Capital \u2013 channels capital toward companies developing on the Innovative Optical and Wireless Network (IOWN) framework.<\/p>\n

This capital deployment mechanism is designed to commercialise a distinct approach to data transport, one built around optical rather than electronic processing, before hyperscalers establish the default standard.<\/p>\n

Deconstructing the IOWN architecture<\/p>\n

The framework rests on three discrete technical pillars, each targeting a specific failure point in existing networks.<\/p>\n

First, the All-Photonics Network (APN). Today\u2019s infrastructure converts data from optical to electronic form at each network node, then back again for onward transmission. That conversion is fast, but it is not free. Each pass consumes power and introduces latency.<\/p>\n

APN keeps data in the optical domain from network core to terminal device\u2014transmitting light, not electrons, along the full path. End-to-end latency drops. So does energy consumption per bit. For AI inference workloads demanding sub-millisecond response times, that difference separates viable deployment from operational failure.<\/p>\n

The second pillar, Digital Twin Computing (DTC), addresses a different constraint. Production-grade digital twins (those a manufacturer runs to simulate an assembly line<\/a> or a logistics operator deploys to model a distribution network) require continuous, bidirectional data exchange between a physical asset and its virtual replica. Today\u2019s networks cannot sustain that exchange without delay. DTC, running over the APN layer, is engineered for that throughput.<\/p>\n

The \u2018Cognitive Foundation\u2019 sits above both layers as an autonomous orchestration engine, allocating network slices, compute, and storage in real-time based on application demand. Self-healing, self-tuning, and requiring no manual intervention. For enterprises deploying AI at the edge, the Cognitive Foundation converts the network from a passive pipe into a programmable service with defined performance characteristics.<\/p>\n

Operational calculus of photonics<\/p>\n

The consortium\u2019s $500 million injection is an acknowledgement of the severe operational challenges facing the digital economy. The exponential growth of AI models and data traffic is placing unsustainable pressure on power grids and existing network infrastructure.<\/p>\n

Current data centres, filled with electronic switches and processors, are becoming a major performance and energy drain. IOWN\u2019s photonics-electronics convergence addresses this. Optical components transmit data at a fraction of the energy consumed by electronic equivalents.<\/p>\n

NTT holds the architectural intellectual property, having built the IOWN Global Forum as the framework\u2019s standards body. SK Group, a South Korean industrial and telecoms conglomerate with its own semiconductor supply chain; and Chunghwa Telecom, Taiwan\u2019s largest carrier, convert what would otherwise be a single operator\u2019s proprietary bet into a potential regional standard.<\/p>\n

These investors bring manufacturing capacity, carrier infrastructure, and procurement scale across two of the world\u2019s largest technology-producing markets. The Development Bank of Japan\u2019s participation adds a further dimension: this is industrial policy dressed as venture capital, positioning IOWN as a priority piece of national infrastructure.<\/p>\n

Enterprise implications beyond bandwidth<\/p>\n

Telesurgery. Remote industrial robotics. Autonomous vehicle command and control. Each of these workloads demands guaranteed round-trip latency in single-digit milliseconds\u2014and APN\u2019s elimination of optical-to-electronic conversion makes such a guarantee technically achievable.<\/p>\n

DTC opens a direct path for industrial enterprises to advance from predictive maintenance to prescriptive optimisation. A manufacturer could run a complete digital replica of its factory floor, feeding in real-time sensor data from hundreds of IoT nodes to simulate proposed production changes before any physical alteration takes place. Fewer unplanned shutdowns. Faster iteration cycles. The network carrying that workload, however, must handle high-volume, continuous data streams without the delays that make today\u2019s cloud-routed twin deployments inadequate for time-sensitive decisions.<\/p>\n

The Cognitive Foundation changes how enterprises procure network services. An organisation deploying AI inference at the edge \u2013 processing data from factory sensors, security systems, or autonomous equipment \u2013 currently has limited paths to guaranteed performance.<\/p>\n

Dedicated leased lines provide a solution but at high cost with limited flexibility. The CF model adds a third option: programmable, autonomous network slices with defined service characteristics, provisioned dynamically without manual intervention. That converts the network from a best-effort utility to a deterministic component of the enterprise IT stack\u2014a level of reliability that public internet connections cannot match.<\/p>\n

Telco-led counterpoint to hyperscaler dominance<\/p>\n

The telecoms industry has watched cloud providers absorb the value layer of the internet economy. The model that took hold during the cloud period placed telcos in the physical transport layer \u2013 bearing capital expenditure, providing capacity, receiving commodity-grade pricing \u2013 while hyperscalers captured economic surplus from services running above the infrastructure.<\/p>\n

AI workloads have deepened that divide. Model training and inference now run predominantly inside the data centres of a handful of technology companies, none of which are telecom operators.<\/p>\n

The IOWN AI Fund is a calculated counter-position. By integrating compute and networking at the architecture level, rather than overlaying compute onto connectivity post-build, the consortium is constructing a value proposition that hyperscalers cannot replicate by purchasing bandwidth alone.<\/p>\n

Distributed AI applications, particularly those requiring real-time local processing, depend on network characteristics that centralised cloud infrastructure cannot deliver. The fund\u2019s mandate is to build out the vendor and startup ecosystem needed to deliver those characteristics commercially.<\/p>\n

Whether the bet pays off depends on adoption timelines and the consortium\u2019s ability to secure developer and enterprise commitment before cloud-native alternatives mature. After all, the $500 million exceeds the annual R&D expenditure of most individual telecom operators outside the top tier.<\/p>\n

See also: BT bolsters network security with Anthropic Claude Mythos AI<\/a><\/p>\n

\"Banner<\/a><\/p>\n

Want to learn more about the IoT from industry leaders? Check out IoT Tech Expo<\/a> taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx<\/a> and is co-located with other leading technology events including AI & Big Data Expo<\/a> and the Cyber Security Expo<\/a>. Click here<\/a> for more information.<\/p>\n

IoT News is powered by TechForge Media<\/a>. Explore other upcoming enterprise technology events and webinars here<\/a>.<\/p>\n","protected":false},"excerpt":{"rendered":"NTT, SK Group, and Chunghwa Telecom have pooled $500 million into a fund for AI infrastructure built on…\n","protected":false},"author":2,"featured_media":48562,"comment_status":"","ping_status":"","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[23],"tags":[1137,9354,15899,28274,8586,26956,11733,8712,16673,16048,4638,958,28275,28276,28277,28278,28279,1130,241,240,8197],"class_list":["post-48561","post","type-post","status-publish","format-standard","has-post-thumbnail","category-sk","tag-ai","tag-autonomous-vehicles","tag-cloud-computing","tag-cognitive-foundation","tag-digital-twin","tag-edge-computing","tag-enterprise-ai","tag-funding","tag-green-tech","tag-industrial-automation","tag-infrastructure","tag-investment","tag-iown","tag-network-slicing","tag-ntt","tag-optical-networks","tag-photonics","tag-robotics","tag-sk","tag-sk-group","tag-sustainability"],"_links":{"self":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/48561","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/comments?post=48561"}],"version-history":[{"count":0,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/posts\/48561\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media\/48562"}],"wp:attachment":[{"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/media?parent=48561"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/categories?post=48561"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.europesays.com\/korea\/wp-json\/wp\/v2\/tags?post=48561"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}