IndiaAI Mission may include Google Tensor chips alongside NVIDIA, AMD
06 Feb 2025

Google may provide its Tensor chip for the next phase of the ₹10,738 crore IndiaAI Mission, according to Moneycontrol.

The move comes as part of the government's continuous efforts to bolster homegrown artificial intelligence (AI) capabilities.

The Tensor chip was first assessed with NVIDIA, AMD, and Intel offerings but failed to receive required legal approvals back then.

IndiaAI Mission's GPU empanelment process
Ongoing evaluation

The GPU empanelment process under the IndiaAI Mission is still underway, with an objective to enhance India's AI compute capacity. Sources indicate that Google's Tensor chip could be part of this effort.

The Tensor chip, a system-on-chip (SoC) processor that powers Google's Pixel smartphones, uses AI to improve performance, security, and camera capabilities.

Projected growth and importance of GPUs
AI expansion

Over the next two months, the number of GPUs procured under the IndiaAI Mission could double from the current 18,693 to some 38,000.

They play a critical role in AI development as they can process multiple tasks at once, much faster than standard computer processors.

Such speed is crucial for training/running AI models that have to process large volumes of data quickly.

Current GPU suppliers and future prospects
Supplier details

In January, the government announced that L1 (lowest 1) bidders and L2 bidders for the August 2024 GPU tender, including firms like Jio Platforms, E2E Networks, and Yotta Data Services will supply GPUs.

Out of the existing 18,693 GPUs, 12,896 are NVIDIA H100 GPUs while 1,480 are NVIDIA H200 GPUs. The rest are high-performance AI chips from AMD, Intel, and AWS.

Potential impact of Google's Tensor chip inclusion
AI optimization

The addition of Google's Tensor chip could expand the GPU lineup and possibly offer a more optimized, cost-effective solution for AI workloads.

This would especially be the case for tasks associated with Google's own AI ecosystem.

The procured GPUs will be offered to start-ups and researchers through a portal that is currently being developed.

Government subsidies and API requirements
Subsidies and APIs

Firms participating in the procurement process have also offered huge discounts on hourly rates.

Further, the government is also offering an additional 40% subsidy for academia, students, and start-ups.

To enable this initiative, application programming interfaces (APIs) will have to be installed at both the suppliers' end and subscribers' end.