Oman advances plans for a 150MW AI data centre
A consortium of companies from Oman, the UAE, and Italy has signed an agreement to develop a new 150MW hyperscale AI data centre in Oman, further reinforcing the sultanate’s growing digital infrastructure ambitions. The proposed “green data centre” will reportedly incorporate renewable energy and high-efficiency systems, aligning with broader efforts to position Oman as a sustainable digital infrastructure hub under Vision 2040. While details around timelines and end users remain limited, the announcement adds to a growing pipeline of digital infrastructure activity across the country, supported by submarine cable investment, international partnerships, and state-backed diversification initiatives. Oman’s strategic position along key subsea cable routes, combined with increasing regional focus on AI infrastructure and connectivity, continues to strengthen its role within the wider GCC digital infrastructure landscape…Read more
SoftBank expands Japan’s sovereign AI infrastructure push
SoftBank Corp. has announced plans to launch a new GPU cloud service for AI workloads in Japan from October 2026, further reinforcing the growing importance of sovereign AI infrastructure and domestic compute ecosystems. The new “AI Data Center GPU Cloud” platform will combine NVIDIA GB200 NVL72 infrastructure with SoftBank’s proprietary Infrinia AI Cloud OS stack, enabling AI training, inference, and large-scale data processing from within Japan-based data centres. The announcement reflects a broader trend emerging globally, where telecom operators, hyperscalers, and infrastructure providers are increasingly moving beyond connectivity and traditional cloud services into vertically integrated AI infrastructure platforms. SoftBank’s strategy also highlights the growing convergence between telecommunications infrastructure, GPU cloud services, and AI-native network architectures, particularly through initiatives such as its Telco AI Cloud and AI-RAN programmes. More broadly, the development underlines how sovereign AI capability is becoming a strategic priority across major economies, with increasing focus on domestic compute availability, operational control, and resilient AI infrastructure stacks capable of supporting long-term AI adoption at scale…Read more
HUMAIN financing plans highlight Saudi Arabia’s AI infrastructure scale
Saudi Arabia-backed AI company HUMAIN has reportedly appointed Goldman Sachs to advise on a financing package worth at least SAR20Bn, or around $5.4Bn, to support new data centre developments and GPU infrastructure in the Kingdom. According to Reuters, the funding would contribute towards 2GW of planned capacity in the Riyadh area, representing roughly one-third of HUMAIN’s targeted capacity by 2034. The report reinforces the scale at which Saudi Arabia is seeking to build out sovereign AI infrastructure, combining domestic data centre capacity, cloud platforms, GPU access, Arabic-language AI models, and enterprise AI services under a PIF-backed platform. It also builds on a series of recent HUMAIN partnerships, including agreements with AWS, AirTrunk, Saudi Aramco, and the National Infrastructure Fund. If realised, the financing would rank among the largest AI infrastructure funding packages in the region, further underlining the Gulf’s growing role in the global AI infrastructure buildout…Read more
Alibaba’s New AI Chip Highlights China’s Push for Compute Sovereignty
Alibaba’s chip unit T-Head has announced a new AI processor, the Zhenwu M890, designed to support both training and inference workloads, further highlighting the strategic importance of domestic chip development within China’s AI infrastructure market. The launch comes as Alibaba’s cloud business continues to report strong AI-led growth, with its self-developed chip capability increasingly positioned as a supply chain advantage in a market shaped by compute scarcity, export controls, and rising demand for GPU and accelerator capacity. The wider trend is also visible at Baidu, where GPU cloud revenue has grown sharply and demand for both training and inference infrastructure remains strong. While Chinese AI chips are still catching up with leading global products in some frontier training use cases, the announcement underlines a broader market direction: AI infrastructure competitiveness is increasingly being shaped not only by data centre capacity and cloud availability, but also by control over the accelerator supply chain itself…Read more
AI Infrastructure Is Reshaping Labour, Talent, and Economic Power
Samsung Electronics’ last-minute agreement with its largest labour union highlights a less discussed consequence of the global AI infrastructure boom: the growing economic and strategic importance of the people building the hardware that powers AI systems. At the centre of the dispute was not declining profitability or industrial contraction, but rather how the immense profits generated by AI-driven demand for memory chips should be distributed across the workforce. As hyperscalers, neoclouds, and AI platforms continue deploying infrastructure at unprecedented scale, semiconductor firms such as Samsung and SK hynix have become some of the most strategically important companies in the global technology stack. The result is a fundamental shift in labour dynamics, where chip engineers and AI infrastructure specialists are increasingly treated as scarce strategic talent comparable to elite software engineers or financial professionals. More broadly, the dispute illustrates how AI is beginning to reshape not only infrastructure markets, but also national economies, labour negotiations, talent competition, wage structures, and industrial policy. As governments and companies race to secure compute capacity, the economic value generated by AI infrastructure is increasingly flowing beyond hyperscalers and into the wider semiconductor ecosystem that enables the AI supply chain itself…Read more