India’s First Blackwell-Powered GPU Cloud Goes Live

Novacore Innovations, based in Mumbai, has launched India’s inaugural GPU cloud platform built around NVIDIA’s Blackwell‑based B200 servers. The platform, powered by a ₹44.6 crore line of credit led by Rashi Fincorp with backing from U. S. and Abu Dhabi investors, is being deployed in Hyderabad to serve startups, researchers and enterprises across India, the U. S. and the Middle East.

This GPU‑as‑a‑service offering centres on the NVIDIA B200, an iteration of the Blackwell architecture that boasts up to 2.3× higher peak performance and double the real‑world AI speed over its predecessors. It is equipped with 192 GB of HBM3e memory, 8 TB/s bandwidth, fifth‑generation tensor cores and dual transformer engines, which collectively accelerate training by up to 3× and inference throughput by as much as 15×, while enhancing energy efficiency up to 25×.

Novacore’s co‑founders, Ranbir Badwal and Aryamaan Singhania, credit India’s low energy costs, technical workforce and Hyderabad’s robust data‑centre ecosystem for enabling this development. They emphasise that the platform offers lower hosting costs, rapid provisioning, and scalable compute for generative AI, large language models, scientific computing and real‑time analytics.

As an incentive to foster uptake, Novacore is offering trial access to qualifying startups and research labs across its target markets.

The platform also aligns with wider developments in India’s AI infrastructure. NVIDIA is collaborating with major domestic firms such as Reliance and Tata to expand locally hosted AI data‑centre capabilities powered by renewable energy, supporting India’s sustainability and net‑zero objectives. Meanwhile, the domestic AI‑as‑a‑service market is forecast to expand at a compound annual growth rate of roughly 26.4 percent, reaching an estimated $31.94 billion by 2031.

The launch of this platform by Novacore marks a significant enhancement in India’s ability to support advanced, high‑performance AI workloads locally. Transactions such as a similar move by CoreWeave in North America—deploying Blackwell Ultra GPUs commercially in July 2025—highlight a global push toward increasingly capable and efficient AI infrastructure.

_{Images are for reference only.Images and contents gathered automatic from google or 3rd party sources.All rights on the images and contents are with their legal original owners.}