Tunisia to Launch AI Campus in Bizerte by Early 2028
SoleCrypt has announced plans to build a dedicated artificial intelligence (AI) campus in Bizerte, Tunisia, with operations slated to begin in early 2028. The project was unveiled at the NVIDIA GTC Europe conference in Berlin.
Infrastructure and Technical Specifications
The facility is designed to reach a total capacity of 40 megawatts (MW), dedicated to the high-performance computing required for developing and deploying AI models.
- Phase 1: The initial phase will cover 7 MW of capacity.
- Hardware: It will feature 1,728 NVIDIA Rubin GPUs, integrated into the NVIDIA HGX Rubin NVL8 platform. These processors are engineered to handle the intensive calculations necessary for training and running large-scale AI models.
Strategic Location and Connectivity
The choice of Bizerte offers significant connectivity advantages. According to SoleCrypt, the site is located approximately 3.6 kilometers from the landing point of the Medusa submarine cable. This proximity is expected to facilitate low-latency connections with European and international networks.
Energy Strategy
Powering AI data centers requires substantial electricity and advanced cooling systems. To support this initiative, SoleCrypt has announced a 60 MW solar power plant with battery storage in Tozeur. The electricity generated will be injected into the Tunisian grid.
However, the specific mechanisms linking the Tozeur solar plant to the Bizerte campus remain to be detailed. The announced solar capacity does not guarantee that the campus will be powered exclusively by this source at all times. Grid connection protocols and energy storage solutions will be critical factors in determining the facility’s energy mix.
Commercial Model and Market Reach
SoleCrypt plans to partner with Hosted.AI to commercialize its computing power as cloud services. The target market includes:
- Enterprises
- Startups
- Research institutions
The primary focus will be on North Africa and the Middle East, with expansion into other international markets. For Tunisian companies, this campus could significantly improve access to advanced computing resources, enabling startups and research bodies to develop their AI applications. Access will, however, depend on the pricing and commercial terms to be established.
Key Partners
The project involves a consortium of specialized partners:
- Schneider Electric: Responsible for power supply, cooling, and facility management.
- Supermicro: Providing expertise in GPU-optimized computing systems.
- Meinhardt and Sterling & Wilson Data Center: Contributing to engineering solutions.
- Maerifa Solutions: Participating in the design of the computing infrastructure.
Outlook and Economic Impact
While the timeline targets early 2028, several key details remain to be finalized, including:
- Total investment amount
- Projected job creation
- Specific access conditions for Tunisian businesses
These factors will be crucial in assessing the campus’s economic impact and its potential to strengthen Tunisia’s position in the regional AI market.
