Indosat Launches Zankore AI Infrastructure Platform in Indonesia
Indosat Ooredoo Hutchison launched Zankore on August 6 with Ooredoo Group, Nokia, and NVIDIA, targeting 1 GW of NVIDIA DSX AI Factory capacity in Indonesia. ANTARA reports that construction is scheduled to start in Batang, Central Java, in the first half of 2027, beginning with roughly 200 MW. The platform is designed to serve Asia-Pacific enterprises, governments, and hyperscalers.
Indosat Ooredoo Hutchison launched Zankore by Indosat on August 6 with Ooredoo Group, Nokia, and NVIDIA, targeting 1 GW of NVIDIA DSX AI Factory capacity in Indonesia. The partners describe the platform as full-stack AI infrastructure for the Asia-Pacific market, with Indonesia as the initial deployment location.
According to Indonesia's ANTARA news agency, construction is scheduled to begin in the first half of 2027 in Batang, Central Java, with initial capacity of approximately 200 MW before expansion toward 1 GW within three years. Communications and Digital Affairs Minister Meutya Hafid told ANTARA that the project could serve domestic and regional demand and strengthen Indonesia's position as a regional AI hub.
Fierce Network reports that Ooredoo Group is providing $800 million in backing for the effort. The full 1 GW target remains a deployment goal rather than installed capacity, and the reports reviewed do not provide a completion date for the final buildout.
Compute, networking, and power management
TechNode Global reports that the initial 200 MW deployment is expected to use NVIDIA GB300 NVL72 systems. It describes NVIDIA DSX as a reference architecture combining accelerated computing, networking, power, cooling, and operational components in a unified AI infrastructure design.
The same report states that the platform is set to incorporate NVIDIA DSX MaxLPS, a power-management capability that dynamically optimizes GPU-fleet power use. NVIDIA claims the technology can enable up to 40% more compute within the same power budget. That figure is a vendor capability claim, not a measured result from the planned Indonesian deployment.
Nokia's August 6 announcement identifies its role as providing AI-native networking, while NVIDIA is contributing accelerated computing, AI software, GPU access, and ecosystem support. Ooredoo Group is identified as lead investor and platform sponsor, and Indosat as the local infrastructure and execution partner.
At gigawatt scale, reliable power delivery, cooling, network fabric design, cluster scheduling, and operational telemetry become as consequential as GPU selection. Across comparable AI data-center projects, the practical value of integrated reference architectures depends on whether power, thermal, and orchestration layers can sustain high accelerator utilization under production workloads.
Sovereign models and AI access
ANTARA reports that Zankore also includes development of Sahabat-AI, an open-source large language model with 70 billion parameters that supports Bahasa Indonesia and regional languages including Javanese, Sundanese, Balinese, and Batak. The platform is also expected to offer Token-as-a-Service access for businesses, government agencies, developers, and the public.
Vikram Sinha, Indosat's president director and CEO, said in a statement reported by TechNode Global that enterprises moving from experimentation to mission-critical AI deployments need an integrated stack spanning models, infrastructure, and network capabilities. Ooredoo Group CEO Aziz Aluthman Fakhroo said in the companies' announcement that AI infrastructure is becoming a fundamental layer of the digital economy.
For ML teams, regional capacity can matter when workloads face data-residency constraints, latency requirements, or limited access to large GPU clusters. However, usable capacity, pricing, model-service availability, developer APIs, and the operating terms for Sahabat-AI have not yet been detailed in the material reviewed.
The announcement places Indonesia in a broader Asia-Pacific competition for sovereign and enterprise AI compute. Fierce Network notes that regional data-center development is expanding amid demand for AI workloads, while power availability remains a constraint for large-scale deployments.
Key Points
- 1Zankore targets 1 GW of AI Factory capacity in Indonesia.
- 2Initial 200 MW capacity is reported for 2027 using GB300 NVL72 systems, making power, cooling, and cluster operations central implementation concerns.
- 3Sahabat-AI and Token-as-a-Service connect infrastructure buildout with local-language model access, though service terms and availability remain undisclosed.
Scoring Rationale
A 1 GW AI-compute target is a major announced infrastructure project for Southeast Asia, with direct relevance to GPU availability, sovereign AI workloads, and regional model deployment. Its score is moderated because the first 200 MW deployment is scheduled for 2027 and the 1 GW figure is a target rather than operational capacity.
Sources
Primary source and supporting public references used for this report.
View 5 more sources
- AI Factory strengthens Indonesia's position as regional AI hub : Govten.antaranews.com
- Ooredoo teams up with Nvidia and Nokia to launch 1 GW AI platform in Indonesiafortune.com
- Indosat-Nokia-NVIDIA Bangun Infrastruktur AI Terbesar di Asia Tenggarafinance.detik.com
- Indonesia's Indosat jointly launches 1GW AI factory platform Zankore with Qatar's Ooredoo, Finland's Nokia, US' NVIDIAtechnode.global
- Indosat launches 1 GW AI infra play with $800M Ooredoo ...fierce-network.com
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
