Firebird has officially launched what it calls the largest NVIDIA DSX AI factory in the region, converting a construction site in Armenia into an operational data center in roughly six months. The company also announced that NVIDIA intends to invest in the business, adding to an earlier investment from CoreWeave, and outlined a global expansion plan targeting 2 gigawatts of AI infrastructure capacity by the end of 2028.
The inauguration event in Hrazdan, Armenia, included remarks from Armenian Prime Minister Nikol Pashinyan, NVIDIA CEO Jensen Huang, and Kazakhstan's Deputy Prime Minister Zhaslan Madiyev, signaling the project's geopolitical weight. The facility is the first operational piece of Firebird's strategy to build AI compute clusters in emerging markets.
Scale and roadmap
Firebird's Armenian AI factory is built on the NVIDIA DSX reference design and uses NVIDIA Spectrum-X Ethernet networking. The company says the design supports up to 40% more GPUs within the same physical footprint. The site is currently operational, and the company plans to scale it to more than 70,000 NVIDIA Rubin and Blackwell GPUs by the end of 2027, reaching 300 megawatts of capacity.
That would place the Armenian platform among Europe's largest AI infrastructure projects. The speed of deployment is a central part of the company's pitch: the facility went from an empty plot to operational status in about six months.
"Just over six months ago, this was an empty construction site. Today, it is the region's largest operational AI factory and the foundation of one of Europe's most ambitious AI infrastructure programs," said Alexander Yesayan, Co-Founder of Firebird. "With a roadmap to scale beyond 70,000 NVIDIA Rubin and Blackwell GPUs, Armenia is rapidly emerging as one of Europe's most important AI computing destinations and a strategic hub for the global AI economy."
Expansion beyond Armenia
Kazakhstan is Firebird's second market. The company has secured 125 megawatts of capacity at Data Center Valley there, with approvals from the Kazakh government and export authorization from the U.S. Department of Commerce's Bureau of Industry and Security. Several additional frontier markets are in development.
NVIDIA's planned investment will support this expansion. The company's stated goal is to reach roughly 2 gigawatts of total capacity across its global pipeline by the end of 2028. For executives tracking AI for Executives & Strategy, the model is notable: Firebird is positioning itself as a turnkey partner for governments that want sovereign AI capacity but lack domestic infrastructure expertise.
"As AI becomes foundational infrastructure for every modern economy, countries need trusted partners who can deliver secure, scalable AI platforms fast," said Razmig Hovaghimian, Co-Founder and CEO of Firebird. "Firebird combines proven execution, deep partnerships with NVIDIA and leading infrastructure providers, U.S. regulatory expertise, and a growing international pipeline. Within two years, we expect our global AI infrastructure platform to reach about 2 GWs of capacity, enabling frontier markets to participate directly in the global AI economy."
Customers and partners
Perplexity is among Firebird's first customers. The AI search company will use the infrastructure to support its answer engine and digital coworker platform. Dell Technologies is providing infrastructure for the Armenian factory; CEO Michael Dell spoke at the launch event.
The project also carries diplomatic significance. Pashinyan noted that the launch comes one year after the U.S.-Armenia memorandum of understanding on AI and technology cooperation, signed with President Donald Trump. NVIDIA's Huang framed the facility in broader terms: "AI factories are the infrastructure nations need to create intelligence, drive economic growth and compete in the age of AI. Together with Firebird, we are building AI infrastructure across Armenia and Kazakhstan that will give researchers, startups and industries the computing foundation to develop AI at home."
For technical teams evaluating infrastructure options, the AI for IT & Development angle is straightforward: the NVIDIA DSX design and Spectrum-X networking are commercially available reference architectures, not proprietary experiments. Firebird's contribution is execution speed and navigating export controls in emerging markets.
Why this matters for executives and strategy
The strategic takeaway is about supply, not technology. Firebird's 2-gigawatt target represents a meaningful addition to global AI compute capacity, and it's being built in countries that previously had no AI infrastructure industry. For enterprises planning AI workloads, this expands the geographic options for data residency and sovereign AI requirements. For governments, it offers a template: partner with a U.S.-backed operator, use NVIDIA's reference design, and get a data center in under a year. Whether that model holds up at scale will depend on Firebird's ability to deliver on its next phases in Kazakhstan and beyond.
Your membership also unlocks:
