Intelligence Journeys
AI Use Cases for Utilities
Private Broadband for Utilities

GPU

August 2026's roundup: EE launches the UK's first commercial 5G slicing service as India's regulator moves first to mandate per-slice tariffs and disclosed speeds, AI-RAN validation reaches SoftBank's live Japanese network, satellite operators from AST SpaceMobile to Elveo Mobile jockey for direct-to-device position, and Charter's completed Cox mega-merger reshapes the US cable map.
August 2026's roundup: Nvidia forms $500B+ AI compute financing platforms with the world's largest asset managers, SK Telecom's SK Hyper data centre unit and Microsoft's new Hyderabad hyperscale region both go live, the custom silicon race widens to include OpenAI's own inference chip and a reported $12.9B Nvidia bid for Hugging Face, and SpaceX's infrastructure ambitions extend from a $16.8B Texas fab to orbital data centers.
August 2026's roundup: operators from SK Telecom to Ooredoo built or monetized AI infrastructure directly, AI-RAN validation reached the device and per-user level, agentic AI moved deeper into live network operations, and governance caught up with a watermarking mandate, an incident-reporting coalition, and a hard look at why AI inference costs keep rising even as token prices fall.
Ooredoo Group and Indosat Ooredoo Hutchison have each launched a branded Zankore GPU-as-a-Service platform in Southeast Asia, targeting 1 gigawatt of AI capacity with Nvidia and Nokia as technical partners. TeckNexus examines the capital structure behind the build, why telecom operators are a plausible GPU-as-a-Service provider in under-served regional markets, and the network-compute bundling trade-off buyers should think through before this kind of vendor relationship reaches their own procurement conversations.
SK Telecom's SK Hyper AI data centre unit, targeting 5 gigawatts of capacity by 2029 and 15 gigawatts by 2035, comes with unusually specific technical architecture detail: a multi-silicon compute platform pairing Arm's AGI CPU with Rebellions' AI inference accelerator, a CXL-based memory pooling architecture built with Panmnesia, and named power and cooling partners in Supermicro and Schneider Electric. TeckNexus examines what each layer reveals, and why this level of disclosed detail is a rare and useful reference point for anyone planning AI infrastructure at any scale.
July 2026's roundup: Hughes Network Systems nears bankruptcy as SpaceX files to launch 100,000 satellites, Nokia's AI-RAN commercial deployments trail Ericsson's 15+ by a full order of magnitude, the FCC finalizes a $6.3B C-band incentive package and closes AT&T's $23B EchoStar spectrum deal, and fiber capital and industrial 5G both keep expanding globally.
May 2026's roundup: the US government's 10% equity stake in Intel (now valued above $50B) underscores how strategic chip manufacturing has become, India and Vietnam both advance new chip fab and testing capacity, the AI chip supply chain keeps fragmenting beyond Nvidia, and Pixxel and Sarvam AI plan India's first orbital data center satellite.

Partner Hubs

Download content, access intelligence tools, and hear from executives.

Partner Events

  • M360 ASEAN
  • FutureNet Asia 2026
  • Network X Vienna 2026
Scroll to Top