Intel and AMD Veterans’ Agrani Labs Seeks $100 Million to Build Next-Generation AI Chips

India’s semiconductor ecosystem could be approaching one of its most significant early-stage deep-tech funding moments yet. Agrani Labs, an artificial intelligence chip startup founded by former Intel and AMD veterans, is reportedly exploring a $100 million Series A funding round as it seeks to accelerate development of next-generation AI chips designed for high-performance computing and large-scale artificial intelligence workloads. Based in Bengaluru, a city increasingly establishing itself as a center for semiconductor research, AI engineering and deep-tech innovation, the company is drawing growing attention at a time when investor appetite around AI infrastructure continues to intensify.

According to reports, discussions remain at an early stage, but if completed, the proposed raise could become one of the largest Series A rounds seen in India’s semiconductor startup ecosystem. Investor conversations are understood to include names such as Qualcomm, Battery Ventures, and other strategic participants with experience across computing and infrastructure technologies. Existing investor Peak XV Partners, which led the company’s seed financing earlier, is also expected to participate and could invest nearly $20 million on a pro-rata basis. While discussions continue, industry observers suggest the scale of the round itself reflects growing investor willingness to place larger bets on deep-tech infrastructure businesses that require longer development cycles and larger capital commitments.

The fundraising effort comes at a time when global competition around AI infrastructure is entering a new phase. Over the last two years, artificial intelligence investment activity has expanded beyond software models and applications into the underlying systems powering them. As businesses increasingly deploy AI across industries ranging from healthcare and finance to manufacturing and enterprise services, demand for computing power has risen sharply. This shift has created a broader race around the hardware enabling artificial intelligence systems to function at scale.

While Nvidia remains the dominant force in AI hardware today, growing demand has created opportunities for newer companies attempting to solve challenges around cost, efficiency and performance. Across global markets, startups and technology companies are increasingly exploring specialized chips optimized for specific AI functions. Investors are also becoming more interested in infrastructure companies capable of serving the next wave of AI growth rather than focusing exclusively on applications sitting above that infrastructure layer.

The company is reportedly developing AI inference chips designed to remain compatible with Nvidia’s CUDA software ecosystem, a move many industry participants view as strategically important. In artificial intelligence, hardware alone rarely determines success. Software ecosystems often become equally important because developers typically build applications around familiar environments and tools. Compatibility with existing systems can reduce friction and potentially accelerate adoption by allowing organizations to integrate new hardware without fundamentally rebuilding existing workflows.

That compatibility decision reflects a larger reality within artificial intelligence markets. For many startups entering AI infrastructure, technological performance alone is no longer enough. Companies increasingly need to convince developers that switching costs remain manageable. Building entirely new ecosystems can be difficult even for well-funded businesses. As a result, newer players are increasingly attempting to align with established standards rather than creating isolated environments.

The focus on inference chips also places the company within one of the fastest-growing segments of AI infrastructure. While model training often attracts public attention because of its scale and visibility, inference is increasingly becoming central to real-world deployment. Inference refers to the stage where trained AI systems process information and generate outputs for users. As AI applications move from experimentation into broader commercial use, demand for efficient inference infrastructure continues expanding rapidly.

Industry analysts suggest this shift could reshape future investment patterns within AI hardware markets. Training large AI systems remains resource-intensive and often concentrated among a relatively small number of technology companies. Inference, however, could ultimately create a broader commercial opportunity because it supports everyday deployment across products and industries. Optimizing infrastructure around this layer may become increasingly important as AI adoption expands.

The development also reflects a broader evolution underway across India’s startup landscape. For years, much of the country's startup activity centered around software services, consumer technology platforms and internet businesses. Semiconductor startups remained comparatively rare because of high development costs, long product cycles and significant technical complexity. Hardware ventures often required larger capital commitments and longer timelines before reaching commercialization.

That environment now appears to be shifting gradually. Increased policy support, growing interest in domestic semiconductor capabilities and rising investor appetite for deep-tech sectors are creating new momentum around advanced hardware technologies. Bengaluru, already home to multiple global semiconductor research and engineering centers, has increasingly emerged as a focal point for this transition. The city’s engineering ecosystem and concentration of technical talent continue positioning it as an important location within India’s evolving semiconductor ambitions.

Earlier this year, the startup emerged from stealth with seed backing from Peak XV Partners, drawing early attention within India’s deep-tech ecosystem. Its focus on AI infrastructure also arrives during a period when global investment activity around semiconductor technologies continues accelerating. Across markets, investors are increasingly looking beyond AI applications toward companies building foundational technologies capable of supporting long-term growth.

The timing of these fundraising discussions is particularly notable because AI infrastructure has become one of the most competitive sectors across global technology markets. Capital is no longer flowing only toward AI models and software companies. Increasingly, attention is shifting toward the chips, systems and computing architectures enabling artificial intelligence to scale.

If completed, the proposed funding round would represent more than a financing milestone for a single startup. It would signal growing confidence that India’s deep-tech ecosystem can increasingly participate in industries historically dominated by global semiconductor leaders and established hardware giants.