Adani Enterprises and US-based manufacturing company Jabil have signed a non-binding agreement to establish a 50:50 joint venture for manufacturing artificial intelligence and data centre equipment in India. The proposed partnership comes as demand for computing infrastructure grows among hyperscalers, cloud providers and data centre operators.

The venture will manufacture and assemble AI racks, servers, storage and networking equipment. Its portfolio will also include power distribution units, liquid-cooling systems and transformers required to operate high-performance computing infrastructure. The companies envisage a multi-gigawatt manufacturing platform serving customers in India and overseas.

The memorandum of understanding remains subject to definitive agreements, regulatory approvals and customary conditions. It follows the companies’ announcement in June of their intention to develop an AI and data centre hardware manufacturing ecosystem in India. The proposed joint venture would also provide installation and commissioning services alongside equipment manufacturing.

The companies said the platform would support India’s expanding data centre market, with industry forecasts indicating capacity of 5-8 GW by 2030. They estimated the global market for AI computing infrastructure at more than USD 3 tn over the next seven years, as investment in computing capacity continues to increase.

The partnership would broaden Adani’s role in the sector from energy supply and data centre development to manufacturing hardware for AI computing. The group has committed USD 100 bn to developing 5 GW of green-energy-powered, AI-ready data centres by 2035. For Jabil, which reported fiscal 2026 revenue of USD 36 bn, the agreement would combine its engineering and manufacturing capabilities with Adani’s energy and infrastructure assets.

The companies said the proposed investment supports India’s plans to expand domestic digital infrastructure and become a manufacturing base for global AI demand. The venture is expected to develop multi-gigawatt AI rack manufacturing capacity, supporting systems with a combined power capacity of several gigawatts.