E2E Networks will use internal accruals and debt to add ~5.4 MW capacity to power NVIDIA’s newest AI chips.

The company told exchanges the plan will support NVIDIA HGX B300 systems and other Blackwell chips, the processors that train and run AI models.

It will not raise fresh equity for the project. MW stands for megawatt, a measure of how much electricity a data centre can draw.

E2E’s existing capacity is about 9.2 MW, so the new addition is more than half that size. No price tag has been given.

The company said the final spend will depend on how much equipment it actually installs.

The work will follow its own rollout plan and the demand for AI computing.

E2E rents out computing power built around these chips to customers who build and run AI tools.

International customers brought in roughly 20% to 25% of its quarterly revenue. The announcement follows the company’s latest results.

On 16 October, E2E reported revenue of Rs 223.8 crore for the September quarter, up 411% from a year earlier.

Profit after tax came in at Rs 72.7 crore, up from Rs 43.9 crore in the June quarter, and its operating profit margin stood at 76.6%.

For the first half of the year, revenue reached Rs 380.5 crore, up 376%.

During that quarter, E2E also signed a binding term sheet worth about Rs 1,000 crore with an Indian sovereign AI company.

It runs until June 2029 and moves an existing cluster of NVIDIA B200 chips from pay-as-you-go to a multi-year commitment.

Another batch of 1,024 B200 chips is due to go live in the October to December quarter, according to management.

In the same period, E2E added VPC Connect to its AI platform, which lets business customers use private, secure networking, and it named Anand Ganapathy as chief revenue officer.

Unlock profitable opportunities every day! Tradz by EquityPandit provides actionable intraday trading signals for stocks and futures. Don’t miss out – download Tradz by EquityPandit and start winning now!