India is continuing to take steps to get AI computing on its shores.
On May 15, 2026, during Prime Minister Narendra Modi’s official state visit to Abu Dhabi, G42 — the Abu Dhabi-based global technology group — and the Government of India formalized the framework and commercial terms for the deployment of Condor Galaxy India, an 8-exaflop AI supercomputing cluster built on 64 Cerebras CS-3 systems. The deal was witnessed by UAE President H.H. Sheikh Mohamed bin Zayed Al Nahyan and PM Modi, and signed between G42 International CEO Mansoor Al Mansoori and India’s Foreign Secretary Vikram Misri.
The announcement builds on a UAE-India digital infrastructure MOU signed in 2024, and marks one of the most consequential AI infrastructure commitments for India to date.
The Companies Involved
G42 is a technology holding group headquartered in Abu Dhabi, with a focus on AI across sectors from molecular biology to space. It operates the Condor Galaxy network — a series of AI supercomputing clusters across the United States — in partnership with Cerebras Systems. This India deployment extends that footprint into one of the world’s largest emerging markets.
Cerebras Systems is the Silicon Valley company behind the Wafer Scale Engine (WSE), the world’s largest AI chip — 56 times larger than the biggest GPU, and claiming over 20x faster inference and training speeds. The company recently completed a notable IPO on Nasdaq. Cerebras CEO Andrew Feldman marked the occasion on X, calling the India deal “the centerpiece of a collaboration between the Governments of the UAE and India,” noting that Cerebras was “honored and humbled.”
C-DAC (Centre for Development of Advanced Computing) is India’s premier government computing research organisation, with decades of experience building indigenous supercomputers under the PARAM series. Under this deal, G42 in partnership with C-DAC will handle installation, deployment, operations, and maintenance of the cluster.
What the Deal Means
At 8 exaflops — that’s 8,000 petaflops — Condor Galaxy India would dwarf India’s current largest AI supercomputer, AIRAWAT, which runs at around 13,170 teraflops. The cluster is designed to power joint R&D in health and genomics, energy, and geospatial analytics, and is positioned as a foundational asset for India’s sovereign AI ambitions.
G42’s Mansoor Al Mansoori described the deployment as converting “energy and compute into sovereign governed nation-scale intelligence” — the kind of language that underscores the geopolitical dimension of this deal. AI supercomputing is no longer just a scientific resource; it’s becoming a measure of national capability.
India’s Broader AI Infrastructure Push
India has been moving fast to build out AI infrastructure, and the Condor Galaxy announcement slots into a crowded field of major commitments.
Google broke ground on a gigawatt-scale AI hub in Visakhapatnam in April 2026, part of a $15 billion five-year investment. Microsoft followed with a $17.5 billion commitment — its largest-ever investment in Asia — to build cloud and AI infrastructure across India. OpenAI has also reportedly been scouting partners for a 1GW datacenter in the country. Even Uber has announced it is setting up its first India data center in partnership with the Adani Group.
The Indian government has been pulling levers to encourage this. It granted data centers infrastructure status in 2022, unlocking low-cost financing. The IndiaAI Mission, launched in 2024 with a $1.2 billion outlay for GPU procurement, is accelerating sovereign compute. And the draft National Data Centre Policy 2025 is proposing tax exemptions for developers for up to 20 years.
Analysts project India’s data centers will reach 2 GW in active capacity in 2026, drawing at least $25 billion in investment that year alone.
The Significance
What makes the Condor Galaxy deal distinct from others in this wave is its sovereign framing. This isn’t a hyperscaler building capacity for global consumption with Indian geography as a cost advantage. It’s explicitly positioned as infrastructure for India’s researchers, institutions, and startups — with C-DAC, one of India’s oldest computing institutions, as operational partner.
For Cerebras, it’s a validation of a bet on non-GPU AI compute architecture at a time when the market is still largely defined by NVIDIA. For G42, it deepens a strategic UAE-India technology relationship that has been building quietly over several years. And for India, it adds another significant piece to a sovereign AI infrastructure puzzle it is assembling with unusual urgency.