NextFin

Peak XV Invests in Indian Startup C2i to Address AI Data Center Power Constraints

Summarized by NextFin AI
  • The global AI race has shifted focus from chip shortages to electrical capacity, with C2i Semiconductors addressing power delivery issues in AI data centers through a $15 million funding round.
  • C2i aims to reduce energy losses by 10% in data centers, which currently waste 15-20% of energy during power conversion, crucial for meeting the demands of the AI boom.
  • Data center electricity consumption is projected to triple by 2035, with power demand expected to rise by 175% by 2030, making energy costs a primary concern for hyperscalers like Microsoft and Google.
  • The success of C2i will depend on its ability to validate its technology with major data center operators, potentially reshaping AI infrastructure distribution and emphasizing the importance of efficient energy use.

NextFin News - As the global race for artificial intelligence supremacy intensifies under the administration of U.S. President Trump, a critical infrastructure bottleneck has shifted from chip shortages to electrical capacity. On February 15, 2026, Peak XV Partners announced it is leading a $15 million Series A funding round for C2i Semiconductors, a Bengaluru-based startup dedicated to solving the power delivery crisis in AI data centers. The round, which brings C2i’s total funding to $19 million, includes participation from Yali Deeptech and TDK Ventures, signaling a strategic pivot by venture capital toward the unglamorous but essential physics of power management.

The problem C2i aims to solve is one of extreme inefficiency. According to Preetam Tadeparthy, co-founder and CTO of C2i, current data center architectures lose between 15% and 20% of their total energy during the conversion process. High-voltage power from the grid must be stepped down thousands of times before it can be utilized by the high-performance GPUs that drive large language models. As data centers scale to meet the demands of the 2026 AI boom, these losses are no longer sustainable. C2i—an acronym for control, conversion, and intelligence—is developing a "grid-to-GPU" platform that integrates power conversion and packaging into a single plug-and-play system, promising to reduce end-to-end power losses by approximately 10%.

This technological shift comes at a time when the macro-economic landscape for data centers is reaching a breaking point. According to a December 2025 report by BloombergNEF, global data center electricity consumption is projected to nearly triple by 2035. Goldman Sachs Research further estimates that by 2030, data center power demand will rise by 175% compared to 2023 levels—an increase equivalent to adding the entire energy consumption of a top-10 industrial nation to the global grid. For hyperscalers such as Microsoft, Google, and Amazon, energy has surpassed hardware as the primary ongoing operational expense. Rajan Anandan, Managing Director at Peak XV, noted that even a 10% to 30% reduction in energy costs could translate into tens of billions of dollars in savings across the industry.

The investment in C2i also highlights the maturing semiconductor ecosystem in India. Anandan compared the current state of Indian deep-tech to the early days of e-commerce in 2008, suggesting that the country is transitioning from a back-office design hub to a source of original, globally competitive hardware IP. This evolution is supported by the Indian government’s design-linked incentives, which have significantly lowered the financial barriers for silicon tape-outs. C2i, founded in 2024 by a team of former Texas Instruments executives including Ram Anant and Vikram Gakhhar, leverages this local talent pool to challenge established power-management giants who have long dominated the data center stack with rigid, component-based architectures.

Looking forward, the success of C2i will depend on its ability to move from fabrication to validation. The company expects its first two silicon designs to return from manufacturing between April and June 2026. Following this, C2i plans to conduct performance trials with major hyperscalers and data center operators in the United States and Taiwan. If the startup can prove its reliability in the high-stakes environment of AI training—where uptime is paramount—it could set a new standard for "system-level" power delivery. This would not only alleviate the immediate power bottleneck but also reshape the geographic distribution of AI infrastructure, allowing facilities to operate in regions where grid capacity was previously considered insufficient.

The broader implication for the tech industry is a return to fundamental engineering. While the previous two years were defined by breakthroughs in software and model parameters, 2026 is becoming the year of the watt. As U.S. President Trump emphasizes domestic infrastructure and energy independence, technologies that maximize the utility of every kilowatt will become strategic assets. C2i’s journey from a Bengaluru lab to the heart of global AI data centers represents a critical test of whether modular, integrated power solutions can keep pace with the insatiable energy appetite of the artificial intelligence revolution.

Explore more exclusive insights at nextfin.ai.

Insights

What are key technical principles behind C2i's grid-to-GPU platform?

What historical factors led to the current power delivery crisis in AI data centers?

What is the current market situation for AI data center power solutions?

What feedback have users provided regarding existing data center power management solutions?

What trends are emerging in the semiconductor industry as it relates to power management?

What recent updates have occurred in government policies affecting India's semiconductor ecosystem?

What are the implications of C2i's funding round for the future of AI data centers?

What challenges does C2i face as it moves from fabrication to validation?

What controversies surround the current state of power management in data centers?

How does C2i's technology compare to established power management solutions in the industry?

What case studies exist that demonstrate the impact of improved power delivery in data centers?

What possible future developments can we expect in AI data center power management?

How might C2i's success influence global AI infrastructure distribution?

What long-term impacts could rising data center power demand have on the tech industry?

What are the main limiting factors for scaling AI data centers sustainably?

How does the Indian semiconductor ecosystem's evolution mirror that of other tech sectors?

What role does venture capital play in advancing power management technologies?

What are the expected energy savings from C2i's integrated power solutions?

Search
NextFinNextFin
NextFin.Al
No Noise, only Signal.
Open App