G42 To Build 8 Exaflop AI Supercomputer In India

UAE-based G42 will deploy an 8 exaflop AI supercomputer cluster in India in partnership with Cerebras, MBZUAI and C-DAC. The system will be nearly 19 times more powerful than India’s current flagship AI supercomputers AIRAWAT and PARAM Siddhi-AI, marking a major leap in the country’s artificial intelligence infrastructure.

India’s artificial intelligence ambitions received a massive boost on the final day of the India AI Impact Summit 2026, as UAE-based technology company G42 signed a term sheet with the Centre for Development of Advanced Computing (C-DAC) to deploy an 8 exaflop AI supercomputer cluster in the country.

The Ministry of External Affairs (MEA) confirmed that the agreement officially initiates the deployment process. Once operational, the cluster will be accessible to startups, research institutions, enterprises, and public sector bodies for AI research, model training, application development, and commercial use.

---

---

What 8 Exaflops Really Means

The proposed 8 exaflop capacity represents a transformational leap in India’s compute ecosystem.

System	Peak Compute Capacity
AIRAWAT	Part of combined 410 petaflops
PARAM Siddhi-AI	Part of combined 410 petaflops
New G42 Cluster	8 Exaflops (≈ 8,000 Petaflops)

With India’s two flagship AI supercomputers AIRAWAT and PARAM Siddhi-AI delivering a combined peak capacity of 410 petaflops, the new cluster will be nearly 19 times more powerful.

This scale positions India among a select group of nations building ultra-large AI compute infrastructure to power next-generation models.

---

---

Strategic Partnerships Behind The Project

The initiative is not a standalone deployment. It is a multi-institution collaboration involving:

Partner	Role
G42	Lead technology partner
Cerebras	AI chip & hardware provider
Mohamed Bin Zayed University of Artificial Intelligence (MBZUAI)	Research collaboration
C-DAC	Implementation & domestic integration

US-based Cerebras, known for building large-scale AI chips optimized for model training, will provide hardware expertise. The inclusion of MBZUAI strengthens the research collaboration angle, while C-DAC anchors the project within India’s sovereign computing framework.

---

---

Why This Is A Big Deal For India’s AI Ecosystem

Compute power has emerged as the most critical bottleneck in AI development. Large language models, generative AI systems, and deep learning frameworks require immense computational resources.

The new cluster aims to:

• Reduce dependency on foreign cloud providers • Support sovereign AI initiatives • Accelerate domestic startup innovation • Enable enterprise-grade AI deployment • Strengthen national research capabilities

Access to high-performance computing has often been limited to global tech giants. Opening this cluster to startups and private entities could democratise AI innovation within India.

---

---

Access Model: Public & Private Sector Participation

According to the MEA, the supercomputer cluster will not be restricted to government use. Instead, it will be accessible for:

Access Category	Use Case
Startups	Model training & AI product development
Academic Institutions	Research & experimentation
Enterprises	Commercial AI deployment
Government Agencies	Policy, defence & public services

This hybrid access model ensures both sovereign control and innovation-driven growth.

---

---

Global AI Race: India’s Positioning

Countries such as the US and China have aggressively expanded AI infrastructure in recent years. Compute capability has become a proxy for technological leadership.

Country	AI Strategy Focus
United States	Private-sector led model scaling
China	State-backed AI infrastructure
UAE	AI-driven economic diversification
India	Sovereign AI + Startup enablement

The collaboration with UAE-based G42 reflects deepening strategic tech ties between India and the Gulf region.

---

---

Economic & Startup Impact

India’s AI startup ecosystem has been growing rapidly, but limited access to high-end compute remains a challenge.

The new infrastructure could:

• Lower model training costs • Attract global AI investments • Enable indigenous foundation models • Strengthen deep-tech research • Improve time-to-market for AI products

For enterprises, this could accelerate automation, predictive analytics, and AI-driven transformation across sectors such as fintech, healthcare, manufacturing, and defence.

---

---

Energy & Infrastructure Considerations

Large AI clusters consume substantial power and cooling resources. Efficient infrastructure planning will be crucial.

Infrastructure Factor	Importance
Power Availability	Critical for uptime
Cooling Systems	High-density compute support
Data Security	Sovereign data protection
Sustainability	Long-term viability

India’s ability to manage energy-efficient AI infrastructure will determine the project’s long-term success.

---

---

FAQs

Q: What is an exaflop?

An exaflop equals one quintillion (10¹⁸) calculations per second a massive measure of computing power.

Q: How does this compare to India’s existing AI systems?

The 8 exaflop cluster is nearly 19 times more powerful than the combined capacity of AIRAWAT and PARAM Siddhi-AI.

Q: Who can access the supercomputer?

Startups, enterprises, academic institutions, and government entities will be able to access it for research and commercial use.

Q: Why is compute power so important in AI?

Advanced AI models require enormous computational resources for training and deployment.

---

---

Final Take

The deployment of an 8 exaflop AI supercomputer by G42 in partnership with C-DAC marks a turning point in India’s artificial intelligence journey. By significantly expanding compute capacity and opening it to both public and private players, India is laying the groundwork for sovereign AI development at global scale.

In the global race for AI dominance, compute is currency and India has just made a powerful investment.

⚠️ DISCLAIMER: This article is for informational purposes only.