E2E Networks Gets INR 177 Cr Order From IndiaAI Mission To Allocate GPUs

AI-focussed cloud service provider E2E Networks has received a contract letter from IndiaAI Mission to provide graphics processing unit (GPU) resources worth INR 177 Cr to AI startup Gnani.ai.

Under this, E2E Networks will be allocating H100 SXM and H200 SXM GPU resources to the startup to enable it develop a foundational model, it said in an exchange filing.

The company said that H200 SXM and H100 SXM GPUs have been allocated for a period of 360 days totalling to 1.30 Cr GPU hours. 

The H100 SXM and H200 SXM are high-performance NVIDIA GPUs built for data centres and large-scale AI workloads to enable enhanced power delivery and faster data transfer. 

The GPUs are to be provided with an InfiniBand (IB) network on a single fabric, the filing added. 

An IB network on a single fabric is a unified, high-performance network infrastructure built using the IB standard. It supports data traffic, including computing and storage, for an entire data centre or a high-performance cluster.

“This engagement further strengthens E2E Networks Limited’s positioning in the national AI infrastructure ecosystem and marks a significant step in supporting India’s ambitions in AI capability building,” the company said.

Its shares surged 10% to hit the upper circuit at INR 2,643 on the NSE today. 

Gnani.ai is one of the four AI platforms selected to build large language models (LLMs) under the IndiaAI mission. While Bengaluru-based GenAI startup Sarvam AI was the first platform to get selected in April to build homegrown sovereign LLM, Soket AI, Gan.ai and Gnani.ai joined it in May. 

Notably, Gnani.ai is building a 14 Bn parameter voice AI foundational model to offer multilingual and real-time speech processing with advanced reasoning capabilities. To support development of the homegrown foundational model, the government is providing compute infrastructure support to the selected startups. 

India’s Ambitious AI Push 

Last month, it was reported that the IndiaAI Mission is set to expand its GPU resources by adding nearly 3,850 GPUs through the third round. IndiaAI Mission’s chief executive Abhishek Singh said that IndiaAI has got 1,050 Google Trillium tensor processing units (TPUs).

India had installed more than 17,300 GPUs under the IndiaAI Mission until June, supported by providers like Hiranandani Group’s Yotta Data Services, E2E Networks, and NxtGen Cloud Technologies.

E2E Networks and NxtGen were the lowest (L1) bidders in the categories where they offered to supply GPUs. While E2E offered to supply 1,353 GPUs, Yotta and NxtGen offered 9,216 and 1,088 GPUs, respectively. 

Besides the aforementioned companies, Jio Platforms and CtrlS Datacenters are also L1 bidders in the tender for sourcing GPUs but they are yet to begin deployment.

The Union cabinet approved the IndiaAI Mission in March 2024 with an outlay of INR 10,372 Cr for a period of five years. IndiaAI Mission aims to build an AI ecosystem, offering supercomputing capabilities and GPUs to various stakeholders, including startups, academic institutions, government agencies and research institutes.