Summary
- Firebird opened an AI factory in Hrazdan, Armenia, setting a target of deploying more than 70,000 NVIDIA GPUs and 300MW of infrastructure capacity by the end of 2027.
- Built on NVIDIA's DSX platform, the facility can house up to 40% more GPUs per unit of floor space, and Firebird has secured investment from both NVIDIA and CoreWeave.
- Perplexity is an early anchor customer, planning to use Firebird's high-performance computing resources for its AI agent platform.
A Major AI Factory Opens in Armenia
Firebird, an emerging AI cloud company, held an opening ceremony for its AI factory in Hrazdan, Armenia on August 8, 2026. The event was attended by Armenian Prime Minister Nikol Pashinyan, Kazakh Deputy Prime Minister Zhaslan Madiyev, and Deputy Chief of Mission David Allen of the U.S. Embassy in Armenia, underscoring the facility's regional significance.
Firebird said it completed the facility in just over six months. The site is built on NVIDIA accelerated computing and Dell Technologies high-performance AI infrastructure, running NVIDIA's reference architecture, accelerated computing, networking, and AI software on top of Dell PowerEdge servers.
Targeting 70,000 GPUs and 300MW by Late 2027
Firebird plans to deploy more than 70,000 NVIDIA Rubin and Blackwell GPUs along with 300 megawatts of AI infrastructure capacity in Armenia by the end of 2027. An AI factory refers to the computing infrastructure required for training, fine-tuning, and deploying AI models, with the core aim of enabling developers, startups, enterprises, universities, and public institutions to build AI tailored to their own languages and industries.
As scale grows, energy efficiency becomes increasingly critical. Firebird's AI factory adopts the NVIDIA DSX platform, integrating accelerated computing, networking, power, and cooling into a single unified system. The company said this design allows up to 40% more GPUs to be operated within the same footprint, producing more tokens per megawatt and greater value per dollar spent.
Investment from NVIDIA and CoreWeave, Plus a 2-Gigawatt Roadmap
Alongside the opening, Firebird announced that NVIDIA plans to invest in the company. Earlier this year, CoreWeave had also invested in Firebird. Both investments are expected to support Firebird's global infrastructure expansion, strengthen its operational capabilities, and help build large-scale computing clusters across emerging markets.
Firebird's expansion plans extend beyond its Armenia site. With NVIDIA's backing, the company is pursuing an AI infrastructure roadmap of roughly 2 gigawatts spanning Armenia, Kazakhstan, and additional markets. Firebird's vision is to position the CIS region as both a hub for global AI companies and a growth platform for local developers, researchers, and businesses.
Partners and Early Demand
Power infrastructure is handled by Schneider Electric, which rapidly delivered and installed core systems—including medium- and low-voltage switchgear, three-phase uninterruptible power supplies (UPS), and rack enclosures—to support Firebird's accelerated deployment timeline. Cooling systems were provided by Vertiv, which built a cooling architecture combining chilled-water technology, advanced control systems, and Vertiv TrimCooler technology, with its iCOM CWM chilled water manager centrally coordinating cooling resources in response to fluctuating AI workloads.
Perplexity has joined as an early anchor customer, saying it is working with Firebird to access the high-performance AI infrastructure needed for its AI agent platform and answer engine. Details such as pricing structure and service region coverage have not yet been disclosed.
Comments