https://aidatacenterindex.com/status/operational/feed/ Chronological feed for Operational using the best public timeline year available for each tracked facility. https://aidatacenterindex.com/datacenters/vantage-quebec-ai-data-center-expansion.html https://aidatacenterindex.com/datacenters/vantage-quebec-ai-data-center-expansion.html Montreal, Quebec, Canada | Operational | 178 MW — Vantage Data Centers has significantly expanded its Canadian footprint in early 2026, reaching a projected total capacity of 178 megawatts (MW) across its Quebec portfolio. This expansion is centered in Montreal and Quebec City, with the co https://aidatacenterindex.com/datacenters/digital-realty-cyberjaya-malaysia.html https://aidatacenterindex.com/datacenters/digital-realty-cyberjaya-malaysia.html Cyberjaya, Cyberjaya, Malaysia | Operational | 14 MW — Digital Realty entered the Malaysian market in early 2026 through the acquisition and expansion of a data center facility in Cyberjaya. The campus currently operates an initial phase (TelcoHub 1) and includes adjacent land for an additional https://aidatacenterindex.com/datacenters/aws-mexico-central-region-queretaro.html https://aidatacenterindex.com/datacenters/aws-mexico-central-region-queretaro.html Querétaro, Querétaro, Mexico | Operational | 345 MW — Launched in January 2025, the AWS Mexico (Central) Region represents a $5 billion USD investment over 15 years. Located in the industrial hub of Querétaro, the region consists of three Availability Zones, each with independent power and coo https://aidatacenterindex.com/datacenters/aws-thailand-cloud-region.html https://aidatacenterindex.com/datacenters/aws-thailand-cloud-region.html Bangkok, Bangkok / Chonburi, Thailand | Operational | 200 MW — Launched in January 2025, the AWS Asia Pacific (Thailand) Region represents a $5 billion USD (approx. 190 billion Baht) investment over 15 years. The region consists of three Availability Zones, each with independent power, cooling, and hig https://aidatacenterindex.com/datacenters/oracle-ais-cloud-thailand-sovereign-region.html https://aidatacenterindex.com/datacenters/oracle-ais-cloud-thailand-sovereign-region.html Bangkok, Bangkok, Thailand | Operational | 120 MW — Launched in July 2025, AIS Cloud powered by Oracle Cloud Infrastructure (OCI) is Thailand's first local sovereign cloud. This partnership between Oracle and AIS involves a $120 million investment in dedicated data center infrastructure with https://aidatacenterindex.com/datacenters/microsoft-chile-central-ai-infrastructure.html https://aidatacenterindex.com/datacenters/microsoft-chile-central-ai-infrastructure.html Santiago, Santiago Metropolitan Region, Chile | Operational | 50 MW — Launched in June 2025, the "Chile Central" region is Microsoft's first data center region in Chile and a major hub for the Southern Cone. Part of the "Transforma Chile" initiative, the region consists of three independent Availability Zones https://aidatacenterindex.com/datacenters/jupiter-exascale-ai-data-center-germany.html https://aidatacenterindex.com/datacenters/jupiter-exascale-ai-data-center-germany.html Jülich, North Rhine-Westphalia, Germany | Operational | 10 MW — JUPITER is Europe's first exascale-class supercomputer, housed in a modular liquid-cooled data center at Forschungszentrum Jülich. Built on Eviden BullSequana XH3000 racks populated with 24,000 NVIDIA GH200 Grace-Hopper Superchips, the syst https://aidatacenterindex.com/datacenters/aist-abci-3-ai-supercomputer-japan.html https://aidatacenterindex.com/datacenters/aist-abci-3-ai-supercomputer-japan.html Tsukuba, Tsukuba, Japan | Operational — Japan's National Institute of Advanced Industrial Science and Technology is rolling out the ABCI-3 platform at its Tsukuba campus to meet surging demand for generative AI. Built on HPE Cray XD systems with NVIDIA H200 GPUs and Quantum-2 Inf https://aidatacenterindex.com/datacenters/aws-project-rainier-ai-campus-indiana.html https://aidatacenterindex.com/datacenters/aws-project-rainier-ai-campus-indiana.html New Carlisle, New Carlisle, Indiana, United States | Operational | 2.4 GW — Amazon Web Services (AWS) is rapidly expanding its "Project Rainier" AI supercomputing campus in New Carlisle, Indiana, representing a total investment of $15 billion. The massive 2.4 gigawatt (GW) campus is specifically optimized to serve https://aidatacenterindex.com/datacenters/xai-gigascale-data-center-tennessee-mississippi.html https://aidatacenterindex.com/datacenters/xai-gigascale-data-center-tennessee-mississippi.html Memphis, Eastern Tennessee / Northern Mississippi, United States | Operational | 1.5 GW — xAI's Memphis Supercluster, nicknamed "Colossus," is one of the most consequential AI infrastructure stories of 2024. Built with breakneck speed by Elon Musk's AI company, the facility assembled 100,000 NVIDIA H100 GPUs in an abandoned semi https://aidatacenterindex.com/datacenters/openai-stargate-data-center-texas.html https://aidatacenterindex.com/datacenters/openai-stargate-data-center-texas.html Abilene, Abilene, Texas, United States | Operational | 1.2 GW — OpenAI's Stargate campus in Abilene, Texas has been built at breakneck pace to meet surging compute demand. The initial phase, live in 2024, houses around 100,000 NVIDIA GPUs drawing roughly 200 megawatts, with construction underway on six https://aidatacenterindex.com/datacenters/google-gemini-cluster-iowa-nebraska.html https://aidatacenterindex.com/datacenters/google-gemini-cluster-iowa-nebraska.html Council Bluffs, Midwest (Iowa–Nebraska cluster), United States | Operational | 1.0 GW — Google has established a massive 1-gigawatt (GW) AI training cluster spanning its campuses in Iowa and Nebraska to serve as the primary engine for its Gemini models. This "AI Hypercomputer" architecture integrates thousands of Google’s cust https://aidatacenterindex.com/datacenters/amazon-nuclear-powered-data-center-acquisition.html https://aidatacenterindex.com/datacenters/amazon-nuclear-powered-data-center-acquisition.html Berwick, Northeastern United States, United States | Operational | 960 MW — In March 2024, Amazon Web Services (AWS) acquired the Cumulus Data Assets data center campus in Berwick, Pennsylvania, for $650 million. The campus is co-located with Talen Energy's 2.5 GW Susquehanna Steam Electric Station, providing AWS w https://aidatacenterindex.com/datacenters/microsoft-japan-ai-data-centers.html https://aidatacenterindex.com/datacenters/microsoft-japan-ai-data-centers.html Tokyo, Tokyo & Osaka, Japan | Operational | 500 MW — Microsoft announced a $2.9 billion (¥440 billion) investment in April 2024 to significantly expand its cloud and AI infrastructure in Japan, its largest-ever investment in the country. This expansion doubles the company’s existing computing https://aidatacenterindex.com/datacenters/microsoft-brazil-sao-paulo-expansion.html https://aidatacenterindex.com/datacenters/microsoft-brazil-sao-paulo-expansion.html São Paulo, São Paulo, Brazil | Operational | 500 MW — In September 2024, Microsoft announced its largest single investment in Brazil, committing R$14.7 billion ($2.7 billion USD) over three years (2024–2027) to expand its cloud and artificial intelligence infrastructure. This expansion focuses https://aidatacenterindex.com/datacenters/microsoft-mexico-cloud-region-queretaro.html https://aidatacenterindex.com/datacenters/microsoft-mexico-cloud-region-queretaro.html Querétaro, Querétaro, Mexico | Operational | 345 MW — In February 2024, Microsoft launched its first cloud data center region in Mexico, officially named "Mexico Central." Located in the state of Querétaro, the region consists of three Availability Zones, each with separate power, cooling, and https://aidatacenterindex.com/datacenters/aws-brazil-cloud-region-expansion.html https://aidatacenterindex.com/datacenters/aws-brazil-cloud-region-expansion.html São Paulo, São Paulo, Brazil | Operational | 300 MW — In September 2024, Amazon Web Services (AWS) announced a 10.1 billion reais (approximately $1.8 billion USD) investment in Brazil over 10 years to expand its cloud and generative AI infrastructure. This new funding focuses on the state of S https://aidatacenterindex.com/datacenters/aws-malaysia-cloud-region.html https://aidatacenterindex.com/datacenters/aws-malaysia-cloud-region.html Cyberjaya, Greater Kuala Lumpur, Malaysia | Operational | 300 MW — Launched in August 2024, the AWS Asia Pacific (Malaysia) Region represents a $6 billion (RM25.5 billion) investment by Amazon Web Services over 15 years. The new region, based in Greater Kuala Lumpur (including major sites in Cyberjaya), co https://aidatacenterindex.com/datacenters/microsoft-malaysia-ai-cloud.html https://aidatacenterindex.com/datacenters/microsoft-malaysia-ai-cloud.html Kuala_Lumpur, Greater Kuala Lumpur / Johor, Malaysia | Operational | 300 MW — In May 2024, Microsoft announced a $2.2 billion investment in Malaysia, its largest in the country's history. The project includes the development of two distinct cloud regions: Malaysia West in Greater Kuala Lumpur and a second hub in Joho https://aidatacenterindex.com/datacenters/apple-ai-private-cloud-compute-us.html https://aidatacenterindex.com/datacenters/apple-ai-private-cloud-compute-us.html Maiden, Multi-site (North Carolina, Oregon, Arizona), United States | Operational | 200 MW — Apple's Private Cloud Compute (PCC) is a purpose-built AI inference infrastructure designed for Apple Intelligence — Apple's on-device and cloud AI system. Unlike general-purpose cloud compute, PCC is engineered around privacy: requests are https://aidatacenterindex.com/datacenters/coreweave-uk-ai-data-centers.html https://aidatacenterindex.com/datacenters/coreweave-uk-ai-data-centers.html London, Greater London and Manchester, United Kingdom | Operational | 150 MW — CoreWeave, a specialized GPU cloud provider focused on AI compute, expanded into Europe with data centers in the United Kingdom starting in 2024. With locations in the Greater London area and Manchester, CoreWeave offers NVIDIA H100 and H20 https://aidatacenterindex.com/datacenters/microsoft-azure-austria-vienna-ai-region.html https://aidatacenterindex.com/datacenters/microsoft-azure-austria-vienna-ai-region.html Vienna, Vienna, Austria | Operational | 120 MW — Microsoft expanded Azure into Austria with a new cloud region in 2024, establishing data residency infrastructure for Austria and extending the DACH (Germany-Austria-Switzerland) coverage footprint alongside existing German regions (West Eu https://aidatacenterindex.com/datacenters/microsoft-azure-colombia-bogota.html https://aidatacenterindex.com/datacenters/microsoft-azure-colombia-bogota.html Bogotá, Cundinamarca, Colombia | Operational | 80 MW — Microsoft opened its first cloud region in Colombia in 2024, anchored by a $150 million infrastructure investment — the largest single cloud infrastructure commitment in Colombian history at the time. The Bogotá region comprises multiple Av https://aidatacenterindex.com/datacenters/google-cloud-africa-johannesburg-region.html https://aidatacenterindex.com/datacenters/google-cloud-africa-johannesburg-region.html Johannesburg, Gauteng, South Africa | Operational | 60 MW — Google Cloud launched its Africa (Johannesburg) region in 2023-2024, joining AWS and Microsoft Azure in establishing hyperscale cloud infrastructure on the African continent. The Johannesburg region reduces latency for South African and bro https://aidatacenterindex.com/datacenters/google-cloud-philippines-manila.html https://aidatacenterindex.com/datacenters/google-cloud-philippines-manila.html Manila, Metro Manila, Philippines | Operational | 60 MW — Google Cloud launched a Philippines region in 2024, bringing hyperscale AI infrastructure to one of Asia's most consequential digital markets. The Philippines is the world's leading destination for Business Process Outsourcing (BPO), genera https://aidatacenterindex.com/datacenters/microsoft-azure-austria-vienna-region.html https://aidatacenterindex.com/datacenters/microsoft-azure-austria-vienna-region.html Vienna, Vienna, Austria | Operational | 60 MW — Microsoft announced the launch of its Azure Austria region in January 2024 with a €1 billion investment commitment. The Austria region is hosted in Vienna and complements Microsoft's existing Germany and Switzerland regions to cover the Ger https://aidatacenterindex.com/datacenters/oracle-cloud-greece-athens.html https://aidatacenterindex.com/datacenters/oracle-cloud-greece-athens.html Athens, Attica, Greece | Operational | 60 MW — Oracle Cloud Infrastructure launched a Greece region in 2024, marking OCI's entry into the Southeastern European market and extending Oracle's EU-resident cloud footprint beyond its existing Germany and Netherlands regions. Greece's EU memb https://aidatacenterindex.com/datacenters/ai71-ai-data-center-uae-abu-dhabi.html https://aidatacenterindex.com/datacenters/ai71-ai-data-center-uae-abu-dhabi.html Abu Dhabi, Abu Dhabi, United Arab Emirates | Operational | 50 MW — AI71 is an Abu Dhabi-based AI company that commercializes the Falcon family of models (developed by TII) for enterprise deployments. The company provides an API platform for Falcon models and customized AI solutions for governments, banks, https://aidatacenterindex.com/datacenters/cyberport-ai-supercomputing-centre-hong-kong.html https://aidatacenterindex.com/datacenters/cyberport-ai-supercomputing-centre-hong-kong.html Cyberport, Hong Kong Island, Hong Kong | Operational | 50 MW — The Cyberport Artificial Intelligence Supercomputing Centre (AISC) is Hong Kong's first large-scale, government-backed AI computing facility. Phase 1 officially commenced operations in December 2024 with a computing power of 1,300 Petaflops https://aidatacenterindex.com/datacenters/kaust-supercomputing-saudi-arabia.html https://aidatacenterindex.com/datacenters/kaust-supercomputing-saudi-arabia.html Thuwal, Makkah, Saudi Arabia | Operational | 35 MW — KAUST's Shaheen III supercomputer, ranked among the world's top 10 most powerful supercomputers at launch in 2024, is powered by HPE Cray EX hardware and uses AMD EPYC processors alongside NVIDIA GPUs. Shaheen III delivers approximately 44. https://aidatacenterindex.com/datacenters/ooredoo-syntys-ai-cloud-data-center-qatar.html https://aidatacenterindex.com/datacenters/ooredoo-syntys-ai-cloud-data-center-qatar.html Doha, Doha, Qatar | Operational | 30 MW — Syn‌tys, Ooredoo Qatar's data center arm, has rolled out the country's first sovereign AI cloud by deploying NVIDIA Hopper GP Us across its Tier III facilities in Doha. The launch equips government agencies, banks, healthcare providers, and https://aidatacenterindex.com/datacenters/egypt-government-data-cloud-computing-center.html https://aidatacenterindex.com/datacenters/egypt-government-data-cloud-computing-center.html Cairo, Cairo, Egypt | Operational | 25 MW — Egypt's Ministry of Communication and Information Technology has opened a 23,500 square meter Tier III data center complex outs ide Cairo to consolidate national databases and unlock AI-enabled public services. Built with contributions from https://aidatacenterindex.com/datacenters/sds-open-access-ai-supercomputer-israel.html https://aidatacenterindex.com/datacenters/sds-open-access-ai-supercomputer-israel.html Petah Tikva, Petah Tikva, Israel | Operational | 18 MW — Shonfeld Data Services has commissioned a ₪1 billion AI supercomputer in Petah Tikva equipped with 2,032 NVIDIA Blackwell B200 GPUs and a high-performance data fabric delivered in partnership with VAST Data. Marketed as Israel's first open- https://aidatacenterindex.com/datacenters/oman-data-park-ai-infrastructure.html https://aidatacenterindex.com/datacenters/oman-data-park-ai-infrastructure.html Muscat, Muscat, Oman | Operational | 10 MW — Oman Data Park has upgraded its Muscat cloud campuses with NVIDIA H200 accelerators, adding to an existing fleet of A100, A40, a nd A10 GPUs to furnish the Sultanate's first domestically managed high-end AI platform. The expansion supports https://aidatacenterindex.com/datacenters/aqaba-digital-hub-mega-data-center-jordan.html https://aidatacenterindex.com/datacenters/aqaba-digital-hub-mega-data-center-jordan.html Aqaba, Aqaba, Jordan | Operational | 6 MW — The Aqaba Digital Hub has opened Jordan's first carrier-neutral Tier III facility with 6 MW of initial IT load, combining coloca tion space, an internet exchange, and high-density pods designed for AI and HPC deployments. Situated at the Re https://aidatacenterindex.com/datacenters/microsoft-australia-ai-expansion.html https://aidatacenterindex.com/datacenters/microsoft-australia-ai-expansion.html Sydney, Melbourne, Canberra, New South Wales, Victoria, ACT, Australia | Operational | 1.0 GW — Microsoft is executing a A$5 billion (US$3.2 billion) investment in Australia, its largest ever in the country, scheduled for completion by 2026. This project focuses on increasing the number of data center sites from 20 to 29 across Sydney https://aidatacenterindex.com/datacenters/oracle-ai-cloud-infrastructure-us.html https://aidatacenterindex.com/datacenters/oracle-ai-cloud-infrastructure-us.html Hillsboro, Hillsboro, Oregon; Chicago, Illinois; Ashburn, Virginia, United States | Operational | 800 MW — Oracle has transformed from a database company into one of the world's largest AI cloud infrastructure providers. In 2024–2025, Oracle committed $50 billion to data center expansion and became the third partner in the Stargate joint venture https://aidatacenterindex.com/datacenters/microsoft-azure-poland-central-warsaw.html https://aidatacenterindex.com/datacenters/microsoft-azure-poland-central-warsaw.html Warsaw, Masovian, Poland | Operational | 500 MW — Microsoft launched the Poland Central region in January 2023 as part of a $1 billion investment in Polish digital infrastructure — its largest single-market investment in Central and Eastern Europe at the time of announcement. The region sp https://aidatacenterindex.com/datacenters/tesla-dojo-supercomputer-new-york.html https://aidatacenterindex.com/datacenters/tesla-dojo-supercomputer-new-york.html Buffalo, New York, United States | Operational | 500 MW — Tesla's Dojo supercomputer at the Gigafactory New York (formerly SolarCity factory) in Buffalo is a purpose-built AI training system designed specifically for training Tesla's Autopilot and Full Self-Driving (FSD) neural networks on video d https://aidatacenterindex.com/datacenters/tencent-tianjin-ai-data-center-china.html https://aidatacenterindex.com/datacenters/tencent-tianjin-ai-data-center-china.html Tianjin, Tianjin, China | Operational | 300 MW — Tencent operates a 300 MW cloud and AI campus in Tianjin that underpins the company's large-model development and nationwide cloud services. The facility uses proprietary high-density liquid cooling to cut energy costs by roughly 30 percent https://aidatacenterindex.com/datacenters/microsoft-azure-italy-milan-region.html https://aidatacenterindex.com/datacenters/microsoft-azure-italy-milan-region.html Milan, Lombardy, Italy | Operational | 150 MW — Microsoft Azure Italy North, launched in 2023, is Microsoft's newest European cloud region and a cornerstone of the company's €4.3 billion commitment to Italian AI and cloud infrastructure announced in 2024. The region offers full Azure AI https://aidatacenterindex.com/datacenters/deepseek-ai-cluster-china.html https://aidatacenterindex.com/datacenters/deepseek-ai-cluster-china.html Hangzhou, Zhejiang, China | Operational | 100 MW — DeepSeek's AI training cluster in Hangzhou, operated by parent company High-Flyer Quant (幻方科技), gained global attention in early 2025 when its DeepSeek-R1 and V3 models matched GPT-4-class performance at a fraction of the training cost — re https://aidatacenterindex.com/datacenters/ibm-watson-ai-us-data-centers.html https://aidatacenterindex.com/datacenters/ibm-watson-ai-us-data-centers.html Raleigh, Multi-site (Virginia, Texas, Georgia), United States | Operational | 100 MW — IBM's watsonx AI platform operates across IBM's existing data center footprint in the United States and internationally, with key US nodes in the Washington DC/Virginia corridor, Texas, and the Research Triangle (NC). watsonx was launched a https://aidatacenterindex.com/datacenters/nebius-ai-cloud-finland-data-center.html https://aidatacenterindex.com/datacenters/nebius-ai-cloud-finland-data-center.html Mäntsälä, Uusimaa, Finland | Operational | 100 MW — NEBIUS (formerly Yandex N.V., renamed after separating international operations from Russian assets) operates a major AI cloud data center in Mäntsälä, Finland — approximately 60 km northeast of Helsinki. The Mäntsälä facility was originall https://aidatacenterindex.com/datacenters/aws-canada-west-calgary-region.html https://aidatacenterindex.com/datacenters/aws-canada-west-calgary-region.html Calgary, Alberta, Canada | Operational | 80 MW — Amazon Web Services launched its Canada West region (ca-west-1) in Calgary, Alberta in November 2023 — becoming the first full AWS region in Western Canada. The launch complemented AWS's existing Canada Central region in Montreal (ca-centra https://aidatacenterindex.com/datacenters/aws-melbourne-ap-southeast-4-australia.html https://aidatacenterindex.com/datacenters/aws-melbourne-ap-southeast-4-australia.html Melbourne, Victoria, Australia | Operational | 80 MW — Amazon Web Services launched its Melbourne, Australia region (ap-southeast-4) in January 2023 — the second full AWS region in Australia, complementing the original Sydney region (ap-southeast-2) that has operated since 2012. The Melbourne r https://aidatacenterindex.com/datacenters/microsoft-azure-portugal-lisbon.html https://aidatacenterindex.com/datacenters/microsoft-azure-portugal-lisbon.html Lisbon, Lisbon, Portugal | Operational | 80 MW — Microsoft launched its Portugal cloud region in Lisbon in October 2022, backed by a €143 million investment commitment. The region was established as part of Microsoft's broader European expansion and provides Portuguese enterprises, govern https://aidatacenterindex.com/datacenters/aws-israel-tel-aviv-il-central-1.html https://aidatacenterindex.com/datacenters/aws-israel-tel-aviv-il-central-1.html Tel Aviv, Tel Aviv District, Israel | Operational | 60 MW — Amazon Web Services launched the AWS Israel (Tel Aviv) region — il-central-1 — in August 2023. The region enables Israeli companies, government agencies, and global enterprises operating in Israel to run cloud workloads with local data resi https://aidatacenterindex.com/datacenters/inflection-ai-midwest-data-center-us.html https://aidatacenterindex.com/datacenters/inflection-ai-midwest-data-center-us.html Chicago, Midwest, United States | Operational | 50 MW — Inflection AI, founded by DeepMind co-founder Mustafa Suleyman and LinkedIn co-founder Reid Hoffman, trained its Pi conversational AI using a high-density GPU cluster at CoreWeave's midwest data center facilities, backed by over $1.3 billio https://aidatacenterindex.com/datacenters/google-cloud-argentina-buenos-aires.html https://aidatacenterindex.com/datacenters/google-cloud-argentina-buenos-aires.html Buenos Aires, Buenos Aires Province, Argentina | Operational | 40 MW — Google Cloud launched its Argentina region (southamerica-west2 / us-east5 subregion context) to serve Argentina's substantial digital economy, including Buenos Aires's outsized concentration of tech talent, fintech companies, and the contin https://aidatacenterindex.com/datacenters/google-cloud-colombia-ai-region.html https://aidatacenterindex.com/datacenters/google-cloud-colombia-ai-region.html Bogotá, Cundinamarca, Colombia | Operational | 40 MW — Google Cloud's Bogotá region, launched in 2023, is the first major public cloud region in the Andean countries of South America (serving Colombia, Peru, Ecuador, Venezuela, and Bolivia). The region offers a low-latency option for Colombian https://aidatacenterindex.com/datacenters/google-cloud-new-zealand-auckland-region.html https://aidatacenterindex.com/datacenters/google-cloud-new-zealand-auckland-region.html Auckland, Auckland, New Zealand | Operational | 40 MW — Google Cloud launched its New Zealand region in Auckland in 2023, providing the first major hyperscale cloud region with full data residency in New Zealand. The region enables New Zealand government agencies (which have strong data sovereig https://aidatacenterindex.com/datacenters/mistral-ai-cluster-france.html https://aidatacenterindex.com/datacenters/mistral-ai-cluster-france.html Paris, Île-de-France, France | Operational | 40 MW — Mistral AI, France's leading AI startup and a major player in the European open-source model ecosystem, trains its models primarily on GPU clusters hosted through OVHcloud (France's largest cloud provider) and other EU-based infrastructure. https://aidatacenterindex.com/datacenters/scaleway-ai-data-center-france.html https://aidatacenterindex.com/datacenters/scaleway-ai-data-center-france.html Vitry-sur-Seine, Paris region, France | Operational | 40 MW — Scaleway, the cloud computing arm of French telecom group Iliad, operates one of Europe's leading GPU cloud platforms, specializing in NVIDIA H100-powered AI compute clusters in its Paris-region data centers. Scaleway is positioned as a Eur https://aidatacenterindex.com/datacenters/cerebras-systems-ai-wafer-cluster-us.html https://aidatacenterindex.com/datacenters/cerebras-systems-ai-wafer-cluster-us.html Dallas, Texas, United States | Operational | 30 MW — The Condor Galaxy AI supercomputer is a joint venture between Cerebras Systems and Abu Dhabi-based G42, designed to provide an alternative to GPU-based AI training. Condor Galaxy uses Cerebras's Wafer Scale Engine (WSE-2) chips — the world' https://aidatacenterindex.com/datacenters/sap-ai-core-germany-data-center.html https://aidatacenterindex.com/datacenters/sap-ai-core-germany-data-center.html Walldorf, Baden-Württemberg, Germany | Operational | 30 MW — SAP runs AI Core infrastructure from its Walldorf, Germany headquarters data centers, powering Joule — SAP's AI assistant embedded across S/4HANA, SuccessFactors, and other enterprise software. SAP AI Core provides the model hosting and inf https://aidatacenterindex.com/datacenters/uk-ai-research-resource-airr-edinburgh.html https://aidatacenterindex.com/datacenters/uk-ai-research-resource-airr-edinburgh.html Bristol / Cambridge, England, United Kingdom | Operational | 20 MW — The UK AI Research Resource (AIRR) is a national initiative funded by the UK government to provide sovereign AI compute for academic research and national AI development programs. The program spans two flagship systems: Isambard-AI at the U https://aidatacenterindex.com/datacenters/marenostrum-5-supercomputing-center-spain.html https://aidatacenterindex.com/datacenters/marenostrum-5-supercomputing-center-spain.html Barcelona, Catalonia, Spain | Operational | 10 MW — MareNostrum 5 is a heterogeneous EuroHPC supercomputer installed at the Barcelona Supercomputing Center. Its Lenovo Neptune liquid-cooled CPU partition spans 6,400 Sapphire Rapids nodes delivering 40 petaflops, complemented by accelerator n https://aidatacenterindex.com/datacenters/airawat-ai-supercomputer-india.html https://aidatacenterindex.com/datacenters/airawat-ai-supercomputer-india.html Pune, Pune, Maharashtra, India | Operational — AIRAWAT—the AI Research Analytics and Knowledge Dissemination Platform—anchors India's national AI infrastructure at C-DAC Pune. Commissioned in 2023, the system delivers 200 AI petaflops (mixed precision) and ranks 75th on the June 2023 To https://aidatacenterindex.com/datacenters/nvidia-israel-1-ai-supercomputer-israel.html https://aidatacenterindex.com/datacenters/nvidia-israel-1-ai-supercomputer-israel.html Israel (cloud access), Israel (undisclosed site), Israel | Operational — NVIDIA is building the Israel-1 AI supercomputer to serve the country's fast-growing ecosystem of more than 800 AI startups. The system, deployed in phases beginning in late 2023, will scale beyond 2,000 NVIDIA H100 GPUs arranged in HGX clu https://aidatacenterindex.com/datacenters/coreweave-ai-data-centers-us.html https://aidatacenterindex.com/datacenters/coreweave-ai-data-centers-us.html Livingston, Livingston, NJ; Plano, TX; Manassas, VA; Chicago, IL, United States | Operational | 500 MW — CoreWeave is the largest independent GPU cloud in the United States, built from the ground up for AI workloads. Founded as a cryptocurrency mining operation in 2017, the company pivoted to NVIDIA GPU compute and became the preferred cloud f https://aidatacenterindex.com/datacenters/oracle-generation-2-cloud-us-ashburn.html https://aidatacenterindex.com/datacenters/oracle-generation-2-cloud-us-ashburn.html Ashburn, Northern Virginia, United States | Operational | 500 MW — Oracle's Ashburn (Manassas) data center campus in Northern Virginia is a core hub of Oracle Cloud Infrastructure (OCI) Generation 2 Cloud, built specifically for GPU-heavy AI and HPC workloads. The region hosts OCI GPU clusters powered by N https://aidatacenterindex.com/datacenters/stc-cloud-saudi-ai-data-center.html https://aidatacenterindex.com/datacenters/stc-cloud-saudi-ai-data-center.html Riyadh, Riyadh, Saudi Arabia | Operational | 200 MW — Saudi Telecom Company's cloud arm, stc Cloud, operates Saudi Arabia's largest national cloud infrastructure, serving as the backbone for the Kingdom's Vision 2030 digital transformation. The Riyadh data centers house AI workloads for Saudi https://aidatacenterindex.com/datacenters/applied-digital-ellendale-north-dakota-gpu-cloud.html https://aidatacenterindex.com/datacenters/applied-digital-ellendale-north-dakota-gpu-cloud.html Ellendale, North Dakota, United States | Operational | 180 MW — Applied Digital Corporation operates one of the largest purpose-built AI cloud data center campuses in North Dakota, located in Ellendale. The facility was originally developed as a high-performance computing (HPC) hosting site for cryptocu https://aidatacenterindex.com/datacenters/aws-zurich-eu-central-2-switzerland.html https://aidatacenterindex.com/datacenters/aws-zurich-eu-central-2-switzerland.html Zurich, Zurich, Switzerland | Operational | 150 MW — AWS opened the Europe (Zurich) region in November 2022, becoming the first major hyperscaler cloud region in Switzerland and a landmark moment for Swiss digital sovereignty. The region spans three Availability Zones across the greater Zuric https://aidatacenterindex.com/datacenters/lambda-labs-ai-cloud-us.html https://aidatacenterindex.com/datacenters/lambda-labs-ai-cloud-us.html Austin, Multi-site (Texas, Washington, Montana), United States | Operational | 150 MW — Lambda Labs operates a distributed AI cloud with GPU clusters across multiple US data centers, providing NVIDIA H100 and A100 compute to AI research labs, startups, and universities. Lambda is the leading alternative GPU cloud to the major https://aidatacenterindex.com/datacenters/aws-hyderabad-ap-south-2-india.html https://aidatacenterindex.com/datacenters/aws-hyderabad-ap-south-2-india.html Hyderabad, Telangana, India | Operational | 120 MW — Amazon Web Services launched its second India region — ap-south-2, based in Hyderabad, Telangana — in November 2022. The Hyderabad region provides a second availability zone cluster within India, complementing the original Mumbai region (ap https://aidatacenterindex.com/datacenters/google-milan-italy-ai-data-center.html https://aidatacenterindex.com/datacenters/google-milan-italy-ai-data-center.html Milan, Lombardy, Italy | Operational | 100 MW — Google Cloud's Milan region (europe-west8), launched in July 2022, is Italy's first hyperscale AI cloud region and serves as the primary Google Cloud access point for Italian enterprises, government agencies, and the broader Southern Europe https://aidatacenterindex.com/datacenters/microsoft-azure-spain-madrid-east.html https://aidatacenterindex.com/datacenters/microsoft-azure-spain-madrid-east.html Madrid, Community of Madrid, Spain | Operational | 100 MW — Microsoft Azure launched its Spain Central region in Madrid in October 2022, backed by a €1.26 billion investment commitment. The region was established to serve Spain's enterprise market with local data residency, supporting compliance wit https://aidatacenterindex.com/datacenters/kakao-ai-data-center-south-korea.html https://aidatacenterindex.com/datacenters/kakao-ai-data-center-south-korea.html Ansan, Chungcheongnam-do, South Korea | Operational | 80 MW — Kakao's cloud data center campus supports KakaoCloud — Korea's third-largest cloud provider — and hosts the AI models powering Kakao's vast consumer platform (KakaoTalk, KakaoMap, Kakao T ride-sharing, and others used by 90%+ of South Korea https://aidatacenterindex.com/datacenters/google-cloud-turkey-istanbul-region.html https://aidatacenterindex.com/datacenters/google-cloud-turkey-istanbul-region.html Istanbul, Istanbul, Turkey | Operational | 50 MW — Google Cloud launched its Turkey region in Istanbul to serve one of the world's fastest-growing digital economies and a strategically positioned country bridging Europe, the Middle East, and Central Asia. Turkey has a population of 85 milli https://aidatacenterindex.com/datacenters/tii-technology-innovation-institute-uae.html https://aidatacenterindex.com/datacenters/tii-technology-innovation-institute-uae.html Abu Dhabi, Abu Dhabi, United Arab Emirates | Operational | 30 MW — The Technology Innovation Institute (TII), Abu Dhabi's advanced scientific research center, developed the Falcon series of large language models — which became the most downloaded open-source AI models in the world in 2023. The Falcon 40B a https://aidatacenterindex.com/datacenters/aleph-alpha-ai-data-center-germany.html https://aidatacenterindex.com/datacenters/aleph-alpha-ai-data-center-germany.html Heidelberg, Baden-Württemberg, Germany | Operational | 20 MW — Aleph Alpha is Germany's leading AI startup, focused on sovereign, explainable AI for European enterprise and government use cases. Its training infrastructure, co-located in German data centers operated with HPE hardware, powers the Lumino https://aidatacenterindex.com/datacenters/lumi-ai-supercomputer-finland.html https://aidatacenterindex.com/datacenters/lumi-ai-supercomputer-finland.html Kajaani, Kainuu, Finland | Operational | 7.1 MW — LUMI resides in a repurposed paper mill in Kajaani, Finland, where abundant hydroelectric power and direct liquid cooling enable one of the world's fastest and greenest supercomputers. The HPE Cray EX system packs 10,240 AMD Instinct MI250X https://aidatacenterindex.com/datacenters/leonardo-ai-supercomputing-facility-italy.html https://aidatacenterindex.com/datacenters/leonardo-ai-supercomputing-facility-italy.html Bologna, Emilia-Romagna, Italy | Operational | 6 MW — Leonardo is a EuroHPC pre-exascale supercomputer hosted at the Bologna Technopole and operated by CINECA. Built on Eviden BullSequana XH2000 technology, the system combines a 3,456-node GPU booster with 13,824 NVIDIA A100 accelerators and a https://aidatacenterindex.com/datacenters/nscc-aspire2a-supercomputer-singapore.html https://aidatacenterindex.com/datacenters/nscc-aspire2a-supercomputer-singapore.html Singapore, Singapore, Singapore | Operational — ASPIRE-2A is NSCC Singapore's flagship supercomputing system, brought online in 2022 to triple the nation's compute capacity versus its predecessor. The platform initially deployed 352 NVIDIA A100 GPUs and has since expanded with NVIDIA H10 https://aidatacenterindex.com/datacenters/sensetime-lingang-ai-data-center-china.html https://aidatacenterindex.com/datacenters/sensetime-lingang-ai-data-center-china.html Shanghai, Lin-gang Special Area, Shanghai, China | Operational — SenseTime brought its Lin-gang Artificial Intelligence Data Center online in January 2022 to power the company's SenseCore supercomputing platform. Built with a ¥5.6 billion investment, the hyperscale cluster delivers more than 2.5 exaFLOPS https://aidatacenterindex.com/datacenters/microsoft-iowa-ai-training-campus.html https://aidatacenterindex.com/datacenters/microsoft-iowa-ai-training-campus.html West Des Moines, Iowa, United States | Operational | 800 MW — Microsoft's West Des Moines campus in Iowa is one of the company's largest data center investments, spanning hundreds of acres with multiple buildings. The Iowa cluster is a key node in Microsoft's Azure AI infrastructure, supporting both g https://aidatacenterindex.com/datacenters/baidu-ai-cloud-data-center-china.html https://aidatacenterindex.com/datacenters/baidu-ai-cloud-data-center-china.html Yanqing, Beijing, China | Operational | 600 MW — Baidu's Yanqi Lake AI Cloud data center in Yanqing District, Beijing, is one of China's largest purpose-built AI training facilities. The facility powers training of Baidu's ERNIE (文心) family of large language models, which compete with GPT https://aidatacenterindex.com/datacenters/volcengine-bytedance-ai-training-china.html https://aidatacenterindex.com/datacenters/volcengine-bytedance-ai-training-china.html Hohhot, Inner Mongolia, China | Operational | 600 MW — ByteDance's Inner Mongolia data center cluster, operated through its cloud arm Volcengine, is one of the largest AI training facilities in China. Inner Mongolia is a preferred location for Chinese hyperscalers due to ultra-cheap coal and wi https://aidatacenterindex.com/datacenters/microsoft-sweden-central-ai-infrastructure.html https://aidatacenterindex.com/datacenters/microsoft-sweden-central-ai-infrastructure.html Gävle, Gävle / Sandviken / Staffanstorp, Sweden | Operational | 500 MW — In June 2024, Microsoft announced a $3.2 billion investment to expand its Sweden Central data center region, specifically to accelerate AI transformation. The expansion includes the deployment of 20,000 advanced GPUs to support high-density https://aidatacenterindex.com/datacenters/huawei-cloud-guizhou-ai-data-center-china.html https://aidatacenterindex.com/datacenters/huawei-cloud-guizhou-ai-data-center-china.html Guiyang, Guizhou, China | Operational | 400 MW — Huawei Cloud's Guizhou data center cluster in Guiyang, the provincial capital of Guizhou, is a major hub for training the Pangu (盘古) large model family, which includes language, vision, and scientific AI models. Guizhou was designated China https://aidatacenterindex.com/datacenters/aws-indonesia-cloud-region.html https://aidatacenterindex.com/datacenters/aws-indonesia-cloud-region.html Jakarta, Jakarta, Indonesia | Operational | 250 MW — The AWS Asia Pacific (Jakarta) Region represents a long-term $5 billion USD investment by Amazon Web Services in Indonesia over 15 years. Since its launch in late 2021, the region has rapidly expanded to meet the soaring demand for AI and c https://aidatacenterindex.com/datacenters/aws-osaka-ap-northeast-3-japan.html https://aidatacenterindex.com/datacenters/aws-osaka-ap-northeast-3-japan.html Osaka, Kansai, Japan | Operational | 150 MW — AWS expanded its Japan presence with the AP Northeast 3 Osaka Region, launched as a full multi-AZ region in March 2021 after operating as a Local Region since 2011. The Osaka region provides geographic redundancy for AWS Japan customers, th https://aidatacenterindex.com/datacenters/google-cloud-indonesia-jakarta.html https://aidatacenterindex.com/datacenters/google-cloud-indonesia-jakarta.html Jakarta, Jakarta, Indonesia | Operational | 100 MW — Google Cloud opened the Jakarta region in 2021, bringing cloud infrastructure to Indonesia — the world's fourth-most-populous country (280 million people), the largest economy in Southeast Asia, and the most important emerging digital marke https://aidatacenterindex.com/datacenters/oracle-cloud-india-ai-region-hyderabad.html https://aidatacenterindex.com/datacenters/oracle-cloud-india-ai-region-hyderabad.html Hyderabad, Telangana, India | Operational | 80 MW — Oracle Cloud Infrastructure (OCI) operates major data center capacity in Hyderabad as one of its two primary India regions (alongside Mumbai). The Hyderabad region hosts OCI compute, storage, and AI/GPU infrastructure serving Indian enterpr https://aidatacenterindex.com/datacenters/google-cloud-chile-santiago-region.html https://aidatacenterindex.com/datacenters/google-cloud-chile-santiago-region.html Santiago, Santiago Metropolitan, Chile | Operational | 50 MW — Google Cloud launched its Chile region (southamerica-west1) in Santiago in April 2021, expanding its South American infrastructure beyond the existing São Paulo region in Brazil. The Santiago region was Google's first cloud region in South https://aidatacenterindex.com/datacenters/meluxina-ai-supercomputer-luxembourg.html https://aidatacenterindex.com/datacenters/meluxina-ai-supercomputer-luxembourg.html Bissen, Luxembourg, Luxembourg | Operational | 3 MW — MeluXina is Luxembourg's national supercomputer, combining AMD EPYC CPU nodes with hundreds of NVIDIA A100 accelerators inside a liquid-cooled BullSequana XH2000 system. Drawing roughly 3 megawatts, the machine achieved a top-five Green500 https://aidatacenterindex.com/datacenters/discoverer-plus-ai-supercomputer-bulgaria.html https://aidatacenterindex.com/datacenters/discoverer-plus-ai-supercomputer-bulgaria.html Sofia, Sofia-Capital, Bulgaria | Operational | 2 MW — Discoverer, Bulgaria's EuroHPC petascale system at Sofia Tech Park, was upgraded in 2025 with four NVIDIA DGX H200 systems tied into its BullSequana XH2000 cluster. The refresh adds 32 H100 GPUs, new high-speed storage, and enhanced UPS cap https://aidatacenterindex.com/datacenters/nvidia-cambridge-1-supercomputer-uk.html https://aidatacenterindex.com/datacenters/nvidia-cambridge-1-supercomputer-uk.html Harlow, England (East of England), United Kingdom | Operational | 1 MW — Cambridge-1 is NVIDIA's national AI supercomputer for the United Kingdom, hosted within Kao Data's DGX-ready campus in Harlow. The system comprises 80 DGX A100 nodes linked with HDR InfiniBand, delivering more than 400 petaflops of AI perfo https://aidatacenterindex.com/datacenters/microsoft-northern-virginia-ai-data-centers.html https://aidatacenterindex.com/datacenters/microsoft-northern-virginia-ai-data-centers.html Ashburn, Northern Virginia (Loudoun, Prince William counties), United States | Operational | 1.0 GW — Northern Virginia — and Ashburn specifically — is the world's densest concentration of data center capacity, and Microsoft Azure anchors a large share of it. Across Loudoun and Prince William counties, Microsoft operates multiple hyperscale https://aidatacenterindex.com/datacenters/alibaba-cloud-zhangbei-ai-data-center-china.html https://aidatacenterindex.com/datacenters/alibaba-cloud-zhangbei-ai-data-center-china.html Zhangbei, Hebei, China | Operational | 500 MW — Alibaba Cloud's Zhangbei Super Data Center sits on the windswept Bashang Plateau in Hebei Province, 260 kilometers northwest of Beijing. The campus is specifically designed for AI training workloads powering the Tongyi Qianwen (通义千问) large https://aidatacenterindex.com/datacenters/ntt-data-ai-singapore-hyperscale.html https://aidatacenterindex.com/datacenters/ntt-data-ai-singapore-hyperscale.html Singapore, Jurong, Singapore | Operational | 250 MW — NTT Data's Singapore data center campus in the Jurong industrial district is one of Southeast Asia's largest hyperscale facilities, serving as a primary hub for AI workloads across the ASEAN region. NTT operates the Singapore facilities as https://aidatacenterindex.com/datacenters/google-fredericia-denmark-ai-data-center.html https://aidatacenterindex.com/datacenters/google-fredericia-denmark-ai-data-center.html Fredericia, South Denmark, Denmark | Operational | 200 MW — Google's Fredericia data center in Taulov is a hyperscale facility that has become a cornerstone of the company's European AI infrastructure. Originally opened in 2020, the site has undergone multiple expansions, including significant land https://aidatacenterindex.com/datacenters/google-south-korea-ai-data-center.html https://aidatacenterindex.com/datacenters/google-south-korea-ai-data-center.html Seoul, Seoul Capital Area, South Korea | Operational | 200 MW — Google Cloud's Seoul region (asia-northeast3) launched in 2020, making South Korea Google's third Asia Pacific region after Tokyo and Singapore. With South Korea ranking among the world's highest AI-adoption economies — home to Samsung, LG, https://aidatacenterindex.com/datacenters/google-cloud-denmark-fredericia.html https://aidatacenterindex.com/datacenters/google-cloud-denmark-fredericia.html Fredericia, South Denmark, Denmark | Operational | 180 MW — Google's Fredericia data center in South Denmark is one of Google's longest-operating European facilities, first opened in 2020 and significantly expanded as AI workloads have grown. Located on the Jutland peninsula at a strategic point nea https://aidatacenterindex.com/datacenters/yandex-cloud-ai-data-centers-russia.html https://aidatacenterindex.com/datacenters/yandex-cloud-ai-data-centers-russia.html Mäntyselkä (Moscow region), Moscow Oblast, Russia | Operational | 150 MW — Yandex, Russia's dominant search engine and internet company, operates AI training infrastructure across multiple data centers in the Moscow region and Siberia. The facilities power YandexGPT — Russia's largest language model — and Alice (А https://aidatacenterindex.com/datacenters/aws-cape-town-africa-south-1-south-africa.html https://aidatacenterindex.com/datacenters/aws-cape-town-africa-south-1-south-africa.html Cape Town, Western Cape, South Africa | Operational | 120 MW — AWS launched the Africa (Cape Town) region in April 2020, becoming the first major hyperscaler to establish a full cloud region on the African continent — a milestone for Africa's digital economy after years of enterprise customers routing https://aidatacenterindex.com/datacenters/deepmind-google-london-ai-research-cluster-uk.html https://aidatacenterindex.com/datacenters/deepmind-google-london-ai-research-cluster-uk.html London, London, United Kingdom | Operational | 80 MW — Google DeepMind operates AI research compute clusters in London, anchored at its headquarters in King's Cross. These clusters support training of frontier models including Gemini (jointly with Google Brain, now merged into DeepMind), AlphaF https://aidatacenterindex.com/datacenters/oracle-cloud-south-korea-seoul.html https://aidatacenterindex.com/datacenters/oracle-cloud-south-korea-seoul.html Seoul, Seoul Capital Area, South Korea | Operational | 80 MW — Oracle Cloud Infrastructure's South Korea region, operating from Seoul since 2020, is OCI's flagship footprint in the Korean market. South Korea's large manufacturing and semiconductor sector relies on Oracle's database-centric AI services