Advanced Micro Devices (AMD) on Monday announced the expansion of its strategic partnership with Tata Consultancy Services (TCS). Under this collaboration, the two companies will build a next-generation rack-scale AI platform based on AMD’s Helios architecture, aimed at accelerating research and innovation across multiple industries.
The initiative will combine AMD’s advanced data-centre processors with TCS’ system integration and services expertise to provide powerful computing infrastructure capable of handling complex workloads.
Rack-Scale AI Platform for Advanced Applications
According to the companies, the partnership will help customers deploy scalable and optimised high-performance computing (HPC) infrastructure that can support next-generation applications. TCS will manage system integration, deployment, and operational services, enabling organisations to upgrade their IT infrastructure and improve efficiency.
AMD said the Helios platform is designed for demanding tasks such as artificial intelligence (AI), machine learning (ML), data analytics, and scientific computing. The platform is powered by AMD Instinct MI455X GPUs, next-generation AMD EPYC “Venice” CPUs, AMD Pensando Vulcano networking cards, and the open ROCm software ecosystem.
Together, these technologies will form a rack-scale AI platform suitable for sovereign AI factories and advanced data-centre environments.
Improved Performance and Flexible Design
The architecture focuses on delivering high compute capability and enhanced networking performance for large-scale AI and machine-learning workloads. AMD Chair and CEO Dr. Lisa Su stated that the Helios platform provides an open AI infrastructure designed for performance, efficiency, and long-term flexibility.
It also supports hybrid cloud environments, advanced networking features, and optimised software stacks for AI and analytics workloads.
200MW AI-Ready Data Centre Blueprint
As part of the partnership, AMD and TCS will also create a blueprint for an AI-ready data centre with capacity of up to 200MW. The design will offer a scalable framework for hyperscalers, enterprises, and AI companies looking to build or expand next-generation facilities.
Additionally, both companies will collaborate with hyperscalers and AI firms to accelerate data-centre development in India.
