AMD and TCS Launch India's First 200MW 'Helios' AI Infrastructure to Support Sovereign Compute Facilities

Mumbai / Santa Clara | EcoPulse24

US-based AMD and India's Tata Consultancy Services (TCS) have expanded their strategic collaboration to launch India’s first rack-scale AI infrastructure platform, 'Helios', with up to 200 megawatts capacity. This project supports India's national AI initiatives and the development of sovereign AI compute factories. The initiative will be executed by HyperVault AI Data Center Limited, TCS’s subsidiary founded in 2025 to build secure, reliable gigawatt-scale infrastructure for cloud computing, AI, and global enterprises.

The new Helios platform features an integrated stack, including AMD Instinct™ MI455X GPUs, next-generation AMD EPYC™ 'Venice' processors, AMD Pensando™ Vulcano NICs, and the open ROCm™ software suite. Helios is designed for high performance, operational efficiency, and long-term flexibility, supporting large-scale deployment of advanced AI training and inference applications.

Under the agreement, AMD and TCS will offer a turnkey blueprint for 'AI-ready' data centers, enabling Indian enterprises to accelerate the transition from pilot projects to large-scale deployment, improving operational efficiency and reducing time-to-market.

Dr. Lisa Su, AMD Chair and CEO, noted that the current AI adoption wave requires a new computing architecture, positioning Helios as an open, high-performance, and future-ready platform. TCS CEO K. Krithivasan highlighted that the partnership lays the foundation for India’s first Helios deployment, combining expertise in AI, connectivity, sustainable energy, and advanced data center engineering.

The 200MW capacity represents an advanced operational scale, supporting thousands of servers dedicated to large language model training and industrial AI applications. The project is part of India’s broader push for digital sovereignty and reducing dependence on foreign data centers, as countries race to build domestic compute capabilities for economic and geopolitical reasons. AMD and TCS will also collaborate with cloud hyperscalers and AI firms to accelerate data center construction within India, strengthening the country's position as a regional advanced computing hub.

A 200MW facility is capable of powering tens of thousands of high-performance GPUs, placing the project among hyperscale AI data centers. HyperVault, founded in 2025, plans to scale AI infrastructure to the gigawatt level, making this project a foundational phase for broader expansion.

India is currently ranked the third most competitive AI nation after the US and China. With growing reliance on local language models and compute infrastructure, governments are seeking to reduce dependence on foreign data centers. The AMD-TCS partnership is aligned with this trend, offering ready-to-deploy infrastructure, a platform for training and inference, reduced deployment timelines, and enhanced operational efficiency.

AMD’s role is evolving from chip supplier to full-stack AI infrastructure provider, offering hardware, networking, software, and architectural design for data centers. This strategic shift reflects the semiconductor industry’s trend toward delivering integrated, turnkey systems, rather than just individual chips.

The move signifies a shift in AI competition from chip-level to integrated infrastructure. For TCS, which posted over $30 billion in consolidated revenue in the fiscal year ending March 2025, the project marks a strategic expansion from digital transformation services into operating large-scale sovereign compute infrastructures.

EcoPulse24 Analysis

The launch of Helios in India with 200MW capacity positions the country at the forefront of AI infrastructure in Asia. If successful in attracting global AI firms, India could become a regional hub for training and operating AI models, with direct impacts on investment, energy, and technology value chains. The initiative is not just technical but has a clear strategic dimension: building national computing capabilities to support a competitive knowledge economy as digital infrastructure control becomes increasingly vital.