BREAKTHROUGH: Bengaluru Startup C2i Just Taped Out India's First AI Power Chip - Peak XV Backs It

· Source: AIM Network · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Cloud Computing & IT Infrastructure, Semiconductor Design & Manufacturing · Depth: Intermediate, short

Summary

Bengaluru-based C2i Semiconductors has taped out India's first AI power chip, a smart power stage chip designed specifically for AI infrastructure. This chip, entirely verified in India, addresses the critical bottleneck of power regulation in increasingly power-intensive AI data centers. Current industry standards involve three power conversion stages, leading to energy loss. C2i's innovation targets the final conversion stage with a software-defined approach, allowing power delivery behavior to adjust via software, enhancing efficiency across server racks. This development comes amid significant global demand for power semiconductors, driven by accelerating data center construction in the US, Europe, India, and Asia. Backed by Peak 15 and founded in 2024 by six semiconductor industry veterans, C2i's achievement marks a significant step for India's semiconductor ambition, shifting from design services for global firms to developing its own AI infrastructure IP.

Key takeaway

For Directors of AI/ML overseeing data center operations, C2i's new software-defined power chip presents a significant opportunity to improve energy efficiency and reliability. You should evaluate this technology as a potential solution for optimizing power delivery in your AI infrastructure, especially given the increasing power demands of scaling AI workloads. This also signals a maturing landscape for indigenous semiconductor innovation.

Key insights

C2i's software-defined power chip addresses AI data center energy efficiency, marking India's first indigenous AI power chip.

Principles

Method

C2i's chip uses a software-defined approach to manage the final power conversion stage, adjusting delivery behavior via software to reduce energy loss in AI data centers.

In practice

Topics

Best for: CTO, VP of Engineering/Data, Executive, AI Hardware Engineer, Director of AI/ML, Investor

Related on AIssential

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AIM Network.