Adani and Jabil Sign 50:50 JV to Build AI Data-Center Hardware at Gigawatt Scale in India
A non-binding MOU signed October 13 pairs Adani's infrastructure ambition with Jabil's contract-manufacturing depth — targeting everything from AI racks to liquid-cooling systems for hyperscalers and cloud operators.
On October 13, 2026, Adani Enterprises and Jabil signed a non-binding memorandum of understanding to form a 50:50 joint venture aimed at manufacturing AI data-center hardware inside India. The deal is non-binding — meaning execution risk is real — but the product scope and scale targets signal something more deliberate than a headline grab.
What the JV Would Actually Build
The planned product portfolio covers nearly every physical layer of a modern AI data center: AI racks, servers, storage systems, networking equipment, power-distribution units, liquid-cooling systems, and transformers. That breadth matters. Most hardware joint ventures carve out a narrow component niche. This one is structured to address the full stack of infrastructure a hyperscaler or cloud operator needs to stand up compute capacity — which either reflects genuine vertical ambition or a negotiating posture designed to attract anchor customers early.
The venture is targeting multi-gigawatt manufacturing capacity for AI infrastructure. That phrase is doing significant work. Gigawatt-scale manufacturing implies not just volume but the kind of sustained, capital-intensive buildout that takes years to reach and requires committed demand on the other side. The stated customer base — hyperscalers, cloud providers, data-center operators, and enterprises — is the right list, but converting MOU language into signed supply agreements is the real test.
Where This Fits Adani's Larger Bet
This JV doesn't exist in isolation. It aligns directly with Adani's stated plan to invest $100 billion in green-energy-powered, hyperscale AI-ready data centers by 2035. That commitment frames the manufacturing venture as a vertical integration play: Adani wants to build the data centers and, increasingly, produce the hardware that goes inside them. Owning both sides of that equation improves margin and reduces supply-chain exposure — particularly relevant as AI hardware demand continues to outstrip the global capacity of established contract manufacturers.
Jabil brings the other half of the equation. As one of the world's largest contract electronics manufacturers, its operational playbook — precision assembly, quality systems, global logistics — is what converts a hardware roadmap into actual shipments. The 50:50 structure suggests neither party is treating this as an outsourcing arrangement. Both sides carry equal governance weight, which typically means both sides expect meaningful revenue from the outcome.
The Harder Questions
The MOU is non-binding, and the distance between an MOU and a functioning gigawatt-scale factory is substantial. A few variables will determine whether this becomes infrastructure or press release.
First, anchor demand. Multi-gigawatt manufacturing capacity requires customers willing to commit volume before the lines are running. Whether Adani's data-center pipeline — and Jabil's existing hyperscaler relationships — can generate that committed demand is unconfirmed.
Second, the green-energy dependency. Adani's $100 billion data-center investment is explicitly tied to green energy. Manufacturing at gigawatt scale carries its own enormous power requirement. The coherence of the full strategy depends on renewable supply chains maturing in parallel with the hardware business.
Third, competitive timing. The global race to stand up AI infrastructure manufacturing outside China is accelerating. India has policy tailwinds, but execution speed matters — a JV that takes three years to reach meaningful output enters a different market than one that ships product in twelve months.
The Shift This Points To
The Adani-Jabil MOU is a data point in a structural reorientation of where AI hardware gets made. For years, the bill-of-materials for a hyperscale data center was assembled across a supply chain concentrated in East Asia, with final integration happening close to the deployment site. What's now emerging — across multiple geographies — is an attempt to build integrated, large-scale hardware manufacturing in regions that also control significant compute demand and energy resources.
India, with Adani's energy and infrastructure reach and a government actively courting semiconductor and hardware investment, is a credible candidate for that role. A 50:50 joint venture combining domestic infrastructure scale with global contract-manufacturing expertise is a structurally sensible vehicle. Whether this particular MOU survives contact with capital allocation, customer commitments, and construction timelines is the question that the next twelve months will begin to answer.