Cerebras Systems is shifting the AI infrastructure conversation from corporate data centers to nation-scale computing projects. The company's wafer-scale engine technology, which debuted publicly through a May 2026 IPO, is now powering Condor Galaxy India, an 8-exaflop AI supercomputing cluster comprising 64 Cerebras CS-3 systems being deployed in partnership with India's government and the Abu Dhabi-based technology group G42.

Why Are Governments Choosing Cerebras Over Traditional GPU Approaches?

The deployment of Condor Galaxy India represents a fundamental shift in how nations approach AI infrastructure. Rather than assembling clusters of graphics processing units (GPUs) from companies like Nvidia, India is betting on Cerebras' alternative architecture. The Wafer-Scale Engine (WSE) processor measures 46,225 square millimeters, making it 58 times larger than Nvidia's B200 chip, and integrates 900,000 compute cores with 44 gigabytes of on-chip SRAM (static random access memory) memory.

The performance advantages are substantial. Cerebras claims its chips can deliver inference speeds up to 15 times faster than leading GPU-based solutions while using only a fraction of the energy. For training frontier-scale AI models, the company reports time reductions exceeding 10 times compared to same-generation GPU systems. These efficiency gains matter enormously when deploying nation-scale infrastructure, where power consumption and operational costs directly impact a country's ability to sustain the project long-term.

"India is one of the world's great innovation economies. Deploying an instance of G42's Intelligence Grid at this scale in such an important geography is what AI-native transformation looks like in practice. We are delivering infrastructure that converts energy and compute into sovereign governed nation-scale intelligence," stated Mansoor Al Mansoori, CEO of G42 International.

Mansoor Al Mansoori, CEO of G42 International

What Makes Cerebras' Wafer-Scale Approach Technically Different?

Cerebras' design philosophy diverges sharply from the industry standard. While most chipmakers have worked to shrink their processors to about the size of a postage stamp, Cerebras went in the opposite direction, designing chips approximately 8.5 inches on a side, roughly the size of a regular iPad. This unconventional approach creates both advantages and manufacturing challenges.

The key technical differences include:

• Memory Architecture: Most GPUs use off-chip high-bandwidth memory (HBM), a special form of DRAM (dynamic random access memory). Cerebras chips instead use on-chip SRAM (static random access memory), which is faster but physically larger and more complex to manufacture.
• Defect Tolerance: Manufacturing a single wafer-sized chip creates a critical vulnerability: if defects appear anywhere on the wafer, the entire processor becomes unusable. To solve this, Cerebras incorporates spare cores into its architecture, enabling chips to work around flawed sections without impacting performance.
• System Integration: Cerebras chips require special cooling and are only sold as part of the CS-3 system, a complete end-to-end server rack solution. The company also offers cloud-based access to these systems.

How Does This Benefit Sovereign AI Development?

The Condor Galaxy India deployment addresses a strategic priority for emerging economies. Under the framework formalized during Prime Minister Narendra Modi's state visit to Abu Dhabi, G42 in partnership with India's Centre for Development of Advanced Computing (C-DAC) will handle installation, deployment, operations, and maintenance of the system. This arrangement gives India direct control over critical AI infrastructure while building domestic expertise.

The supercomputer will underpin research across multiple sectors including health and genomics, energy, and geospatial analytics. Researchers, institutions from both countries, and India's emerging innovators will gain access to frontier-scale computing capacity to address consequential challenges. This represents a departure from the traditional model where nations rely on cloud services from US-based technology companies.

What's the Commercial Context Behind Cerebras' Growth?

Cerebras' IPO in May 2026 reflected strong market confidence in the AI infrastructure sector. The company's stock opened at $350 per share on its first trading day, compared to the IPO price of $185, representing a 68% surge. The company sold approximately 28 million shares of Class A Common Stock at $120 per share during the IPO.

The company's revenue trajectory supports investor enthusiasm. Last year, Cerebras grew revenue by 76% to $510 million with a gross margin of 39%, with approximately 70% of revenue coming from hardware and the remainder from cloud and other services. More significantly, the company reportedly has $20 billion in commitments from OpenAI, suggesting substantial future revenue growth.

The India deployment extends Cerebras' existing partnership with G42. The two companies already operate several clusters of supercomputing capacity across the United States through the Condor Galaxy network, with the India deployment extending that footprint into one of the world's most consequential emerging markets.

Steps to Understanding Cerebras' Role in Global AI Infrastructure

• Recognize the Architecture Shift: Cerebras represents a fundamentally different approach to AI computing compared to GPU-based systems. Rather than assembling many small processors, the company builds single, massive chips with integrated memory and cooling, changing how data centers are designed and operated.
• Understand the Sovereign Angle: Governments are increasingly viewing AI infrastructure as strategic national assets. Cerebras' approach appeals to nations seeking to build independent computing capacity rather than relying on cloud services from US technology companies, as evidenced by India's partnership with G42.
• Track the Manufacturing Challenge: Cerebras' success depends on Taiwan Semiconductor Manufacturing's ability to produce wafer-scale chips with minimal defects. This single-source dependency creates both opportunity and risk for the company's long-term growth trajectory.

The Condor Galaxy India project signals a broader trend: as AI becomes increasingly central to national competitiveness, governments are willing to invest in alternative infrastructure approaches that offer greater control and efficiency, even if those approaches differ from the GPU-dominated status quo. Cerebras' technology, once dismissed as a niche player, is now powering some of the world's most ambitious AI infrastructure initiatives.