Data Centres • Cloud Infrastructure • Digital Sovereignty
Published by GreyRadius Consulting | 3–9 August 2026
Executive Highlights
IndonesiaIndosat launches Zankore with a target of up to 1GW of AI data-centre capacity.
IndiaMicrosoft opens its fourth India cloud region with three Availability Zones and US$20.5B of announced cloud/AI commitments.
EuropePartners Group commits more than US$1B of equity to AVK as grid connections in FLAP-D can take 7–15 years.
SpainAsturias backs a US$1.42B data-centre project targeted for operation by 2030.
Southeast AsiaCoreWeave enters APAC with three Indonesian facilities totaling 360MW of contracted IT power.
United StatesNSF launches regional AI infrastructure hubs with a reported US$100M funding envelope.
EuropeVolta Infra announces a US$10B AI-cloud agreement alongside a US$5B infrastructure initiative.
AI Infrastructure | Data Centres | Sovereign Compute
Indosat targets up to 1GW of AI data-centre capacity through Zankore
7 Aug 2026 | Indonesia | Indosat / Ooredoo Group / NVIDIA / Nokia
Indosat Ooredoo Hutchison, together with Ooredoo Group, NVIDIA and Nokia, launched Zankore by Indosat with a target of up to 1GW of AI data-centre capacity. The ambition is unusually large for a new national AI-compute platform and comes alongside CoreWeave’s separate 360MW Indonesian commitment, taking the two announced capacity signals highlighted this week to as much as 1.36GW. The combination of telecom infrastructure, accelerated-computing technology and data-centre capacity gives Indonesia a potential path from regional hosting market to large-scale AI-compute hub.
The decisive metric is no longer the 1GW headline—it is the conversion rate from target capacity to secured sites, utility power, financing and anchor workloads. Readers should watch whether Zankore discloses phased MW deployment, power sources and customer commitments. Those milestones will show whether Indonesia can turn sovereign-AI ambition into bankable, commissioned infrastructure before competing Asian markets absorb the same capital and GPU demand.
GreyRadius Insight
Indonesia is beginning to assemble an AI-infrastructure ecosystem rather than isolated data-centre projects. If telecom distribution, NVIDIA-class compute and gigawatt-scale capacity are integrated successfully, the country could capture more of the AI value chain locally—from model training and inference to sovereign workloads. For investors and operators, the opportunity is increasingly in the enabling layer: power, cooling, connectivity, land and local enterprise demand—not simply rack capacity.
Cloud Infrastructure | Digital Sovereignty | Hyperscale Cloud
Microsoft opens India South Central cloud region in Hyderabad
6 Aug 2026 | India | Microsoft
Microsoft put its Hyderabad-based India South Central region into general availability with three Availability Zones, making it the company’s fourth cloud region in India alongside Pune, Chennai and Mumbai. The expansion sits behind US$20.5B of announced India cloud and AI investment commitments—US$3B announced in January 2025 plus US$17.5B in December 2025. Microsoft also reports a global footprint of more than 80 regions, 500 data centres and 800,000km of terrestrial and subsea fibre, giving the Hyderabad launch both domestic-residency relevance and global-network scale.
The strategic test is workload movement, not region count. Watch whether regulated enterprises redesign architectures around four domestic regions for residency, low-latency operations and cross-region disaster recovery—and whether Microsoft’s US$20.5B commitment produces visible additions in data-centre capacity, AI compute and connectivity. Adoption by banks, government-linked entities and large enterprises would be a stronger demand signal than infrastructure commissioning alone.
GreyRadius Insight
India is moving toward a multi-region cloud architecture in which resilience and sovereignty can be designed domestically rather than through offshore failover. That changes the enterprise cloud decision: location becomes less of a constraint, while architecture, governance and workload economics become more important. Providers that can combine domestic redundancy with AI capacity and compliance tooling should gain an advantage in regulated workloads.
Data Centre Power | Resilience | Infrastructure Investment
Partners Group commits more than US$1B to AVK data-centre power platform
6 Aug 2026 | Europe | Partners Group / AVK
Partners Group agreed to acquire a majority stake in AVK and plans an initial equity investment of more than US$1B, supplemented by debt. AVK has installed approximately 3.5GW of power across more than 20,000 projects and has an advanced pipeline exceeding 2GW of behind-the-meter power opportunities. Partners Group says grid connections in major FLAP-D markets can take 7–15 years—an infrastructure bottleneck that materially increases the value of alternatives capable of bringing reliable power to data-centre sites faster.
Power-delivery lead time is becoming as important as land acquisition or compute procurement. Watch how much of AVK’s 2GW+ pipeline reaches financial close and commissioning, and whether customers use behind-the-meter systems as bridges to grid connection or as permanent architectures. The investable signal is the number of months removed from time-to-power, because every avoided year can pull forward data-centre revenue.
GreyRadius Insight
The infrastructure value pool is shifting upstream from servers and buildings toward electricity orchestration. When grid queues stretch to 7–15 years, a data-centre site without a credible power pathway is effectively stranded capacity. This creates a strategic premium for platforms that can package generation, storage, backup and grid interaction into a financeable power solution—and makes energy capability a core competitive advantage for developers.
Data Centre Investment | Regional Expansion | Infrastructure Development
Asturias backs US$1.42B data-centre project in Salas
5 Aug 2026 | Spain | Digital Valley Asturias
The Digital Valley Asturias project in Salas envisages approximately US$1.42B of investment, converting the announced US$1.42B at the 5 August 2026 EUR/USD rate. Officials said the project could create 120 direct jobs and nearly 1,000 indirect jobs, while construction could involve up to 3,000 workers. The site is targeted to become operational by 2030, positioning Asturias as another European region seeking to attract data-centre investment beyond the established FLAP-D hubs.
The project should be judged against four conversion gates: permitted land, contracted power, financed construction and committed customers. The 2030 operating target leaves meaningful execution risk, so announcements on grid capacity and anchor tenants will matter more than headline investment. Watch whether Asturias can demonstrate a structurally better time-to-power and cost base than congested European hubs.
GreyRadius Insight
Europe’s next data-centre winners may be selected by infrastructure availability rather than proximity to legacy hubs. Secondary regions can compete if they combine power access, suitable land, fibre routes and permitting certainty. For regional governments, the lesson is equally important: incentives alone are insufficient; the product being sold to operators is an execution-ready infrastructure corridor.
AI Infrastructure | AI Cloud | Data Centres
CoreWeave enters Asia-Pacific with three Indonesian data centres
4 Aug 2026 | Indonesia / Southeast Asia | CoreWeave
CoreWeave announced three Indonesian data centres totaling 360MW of contracted IT power, scheduled to come online in 2028—its first physical data-centre footprint in Asia-Pacific. As of 31 March 2026, CoreWeave operated 49 data centres, more than 1GW of active power and more than 3.5GW contracted globally. The Indonesian commitment therefore equals roughly 10% of its reported contracted global power base and establishes a substantial first step into APAC rather than a small edge deployment.
Watch whether the 360MW programme arrives as synchronized power, buildings and GPU clusters by 2028. For AI infrastructure, contracted MW only becomes commercially meaningful when it is energized and matched with accelerators and customers. Anchor workload announcements, GPU procurement and utility milestones will reveal whether Indonesia becomes CoreWeave’s APAC platform for regional demand or primarily a capacity location.
GreyRadius Insight
Neocloud expansion is creating a new geography of AI infrastructure that does not have to mirror traditional cloud-region maps. A 360MW first entry suggests operators are selecting markets around power availability, scale and future AI demand rather than only existing enterprise-cloud concentration. That can redirect infrastructure investment toward markets able to offer large, executable power blocks.
Government AI Infrastructure | Research Compute | Public Policy
NSF launches State and Regional AI Infrastructure Hubs initiative
4 Aug 2026 | United States | National Science Foundation
The U.S. National Science Foundation launched a State and Regional AI Infrastructure Hubs initiative intended to expand compute access for researchers, students and educators. DatacenterDynamics reported a US$100M funding envelope. NSF funding is aimed at hub coordination, AI-infrastructure workforce development and faculty training, while participating consortia are expected to finance new or expanded compute infrastructure—using federal coordination funding to catalyse a broader regional investment base.
The multiplier matters more than the US$100M headline. Watch how much state, university and private capital each federal dollar attracts, where hubs are located, and whether compute becomes persistently accessible rather than allocated through short-term programmes. A strong outcome would be durable regional capacity plus trained operators and researchers, not simply new hardware purchases.
GreyRadius Insight
Public AI policy is moving from model research toward infrastructure access. Regional hubs can reduce the concentration of advanced compute in a small number of hyperscalers and elite institutions, but their strategic value depends on utilization, talent and sustainable operating economics. The emerging policy question is therefore not only who funds GPUs, but who can operate shared AI infrastructure efficiently at scale.
AI Cloud | Infrastructure Investment | Strategic Partnerships
Volta Infra announces US$10B AI-cloud partnership and US$5B infrastructure initiative
4 Aug 2026 | Europe | Volta Infra / Bitdeer / Azora
Volta Infra announced a US$10B cloud-computing agreement with an undisclosed AI company, working with Bitdeer, alongside a separate US$5B AI-infrastructure initiative with Azora. Reuters reported Volta at a US$2.4B valuation. Together, the announced agreements represent US$15B of headline commercial and infrastructure commitments—more than six times the reported valuation—illustrating the extraordinary contract and capital scale now surrounding AI-compute platforms.
Headline contract value should be separated from deployable economics. Watch counterparty disclosure, contract duration, minimum purchase commitments, financing structure, MW/GPU deployment and revenue-recognition timing. The key diligence question is how much of the US$15B headline value is firm, funded and linked to specific powered capacity; that determines whether the agreements genuinely de-risk infrastructure build-out.
GreyRadius Insight
AI infrastructure is increasingly being financed through a demand-first model: secure long-duration compute commitments, then use contracted demand to support capital formation and capacity build-out. This can accelerate expansion, but it also concentrates counterparty and execution risk. For investors, contracted backlog quality—not simply its size—should become a central metric alongside power availability and cost per delivered unit of compute.
Market Data & Intelligence
Metric
Latest Value
Reporting Period
Region
Executive Implication
Indosat Zankore target
Up to 1GW
7 Aug 2026
Indonesia
Signals gigawatt-scale sovereign and regional AI-compute ambition.
Microsoft India cloud footprint
4 regions / 3 AZs in new region
6 Aug 2026
India
Expands domestic cloud residency and resilience options.
Microsoft India investment commitments
US$20.5B
6 Aug 2026
India
Links new region capacity to a large multi-year cloud and AI investment programme.
AVK installed power / pipeline
~3.5GW installed / >2GW pipeline
6 Aug 2026
Europe
Shows power infrastructure becoming a direct data-centre investment category.
Digital Valley Asturias investment
US$1.42B
5 Aug 2026
Spain
Supports data-centre expansion beyond Europe’s traditional core hubs.
CoreWeave Indonesia contracted IT power
360MW across 3 facilities
4 Aug 2026
Indonesia
Establishes CoreWeave’s first physical APAC data-centre footprint.
NSF AI Infrastructure Hubs funding
US$100M reported
4 Aug 2026
United States
Broadens regional access to research and education compute infrastructure.
Volta Infra announced agreements
US$10B cloud + US$5B infrastructure
4 Aug 2026
Europe
Highlights scale of private AI-cloud contracting and infrastructure capital.
Turn Infrastructure Strategy Into Execution
GreyRadius helps leadership teams translate cloud, data centre and AI infrastructure shifts into market priorities, investment choices and execution-ready growth plans.
