en.Wedoany.com Reported - Indian data center operator Yotta Data Services has completed a funding round of approximately $150 million, valuing the company at around $3.9 billion, equivalent to 370 billion Indian rupees. The funds came from non-institutional investors, with no promoters selling shares. All capital has been injected into the company's operations to support AI and cloud business expansion.
This funding round directly serves Yotta's GPU computing power expansion. The company plans to expand its AI cloud infrastructure to over 40,000 NVIDIA Blackwell GPUs within the next four months, and further increase it to approximately 85,000 GPUs by the end of fiscal year 2027. Blackwell GPUs are primarily designed for large model training, inference, enterprise AI applications, generative AI platforms, and high-performance computing tasks. When deployed in data centers, they do not operate as standalone graphics cards but must be integrated with servers, liquid or air cooling systems, high-speed networks, storage clusters, power systems, scheduling platforms, and cloud service software to form an AI computing infrastructure. For Yotta, the number of GPUs is merely an external indicator of computing scale; the more critical factor is whether it can create sellable, schedulable, isolatable, and auditable AI cloud resources, enabling enterprise clients, government agencies, and developers to access training and inference capabilities via cloud services.
This funding was not achieved through promoter stake reductions, with capital flowing directly into company operations. This arrangement allows Yotta to allocate funds for equipment procurement, cluster deployment, data center expansion, cloud platform development, and customer delivery, rather than merely completing equity-level valuation adjustments.
The Indian AI cloud market is rapidly amplifying local computing power demand. Large enterprises require localized model training and inference resources, governments and financial institutions focus on data residency, compliance audits, and sovereign cloud environments, while software service providers need sustainably accessible GPU resources to serve global clients. Yotta's emphasis on NVIDIA Blackwell GPUs in this expansion indicates its goal extends beyond standard cloud hosts or traditional colocation services, targeting the provision of underlying computing power for AI model companies, enterprise AI applications, developer platforms, and high-density inference scenarios.
After expanding the GPU cluster to over 40,000 units, Yotta must simultaneously address issues related to power, cooling, networking, and resource scheduling. AI servers demand high power supply stability and data center cooling density, while Blackwell-level clusters require high-speed interconnectivity to support multi-node, multi-GPU training and inference tasks. If networking, storage, or cooling capabilities lag behind, merely increasing the number of GPUs will not generate effective computing power. Sovereign cloud services also require that customer data, model operations, access permissions, and audit logs remain within a closed local environment, which cannot be sustained by hardware scale alone.
Yotta's current expansion pace will push India's AI infrastructure toward larger-scale deployment. The goal of exceeding 40,000 Blackwell GPUs within the next four months determines the speed of short-term equipment arrival, racking, networking, and cloud platform integration. The target of approximately 85,000 GPUs by the end of fiscal year 2027 requires the company to continue expanding in data center capacity, energy supply, customer contracts, and operational systems.
