US-based CoreWeave started publicly trading on the American stock exchange Nasdaq last year in March. By raising $1.5 billion in its share sale, at the time it became the biggest tech IPO in the US since 2021. The company is part of a wave of large infrastructure players now being called neoclouds.
Broadly speaking, neocloud companies are cloud providers that specialise in supporting AI workloads by offering compute infrastructure in the form of GPU-as-a-service (GPUaaS). In India too, the neocloud market is seeing significant growth.
Players like Yotta Data Services, NxtGen, NeevCloud, and E2E Networks are some of the noteworthy names that are vying for the neocloud opportunity.
GPUs or graphic processing units are central to the AI revolution with foundational and large language models (LLMs) and AI applications all requiring GPU compute power and high bandwidth memory.
The opportunity is ripe for neocloud companies as the world faces disruption from AI, growing demand-supply gap for GPUs, supply chain bottlenecks, and cost efficiency as compared to traditional hyperscalers.
Neoclouds started appearing more prominently between 2023-2024. According to realty market research group JLL, the neocloud segment has seen a five-year compound annual growth rate (CAGR) in revenue of 82% since 2021.
For over a decade, hyperscalers have been central to India’s cloud adoption, offering elastic compute largely optimised for IT and tech workloads. But as AI moved from experimentation to real-world deployment, cracks began to appear in the form of advanced GPU scarcity, often considered the foundational backbone.
According to industry executives, hyperscalers, constrained by capex cycles and limited GPU supply, have increasingly rationed advanced capacity. Neoclouds help bridge a gap by providing accessible, scalable, and local AI compute, especially for startups, deep-tech firms, research institutions, and enterprises that cannot always rely on hyperscalers for cost, availability, or sovereignty reasons.
The New Hyperscalers
“Neoclouds could realistically capture 5–10% of the overall cloud market and as much as 30% of AI/GPU workloads over the next 3-7 years,” said Seema Ambastha, CEO of Larsen & Toubro’s data centre business Vyoma.
Ambastha also pegged neoclouds to capture 10–15% of cloud infrastructure and 25–35% of AI compute by 2030 in India.
Notably, L&T Vyoma has begun work on its latest Navi Mumbai data centre — “a 100 MW campus with purpose-built to support build-to-suit, enterprise and neocloud workloads”, as the CEO puts it — looking to cater to demand and workloads from AI-native companies, enterprises in banking and finance, retail, healthcare, and government.
In several cases, an industry shift to hybrid and multi cloud environments is also shaping neocloud adoption.
“Large hyperscalers already have massive enterprise customer bases, which makes it difficult for neocloud players to displace them for public cloud workloads. However, enterprises will increasingly adopt hybrid and multi-cloud strategies, combining public cloud with on-premise, colocation, and edge deployments,”said a Bengaluru-based industry analyst.
In their opinion, neocloud providers are well positioned to grab market share for regulated, latency-sensitive, and compliance-heavy workloads, particularly in BFSI, manufacturing, and mobility.
India’s Neocloud Opportunity
India presents a particularly strong case for neocloud adoption. AI demand across startups, enterprises, academia, and government programs is rising sharply. Neocloud demand is driven by local data needs, sovereign AI initiatives, and cost-sensitive scaling requirements. In terms of cost, industry estimates suggest that neoclouds offer services at 20-30% less than traditional hyperscalers.
This shift is also being reinforced by public-sector demand, including large national programs such IndiaAI Mission, focused on building AI capacity at scale.
India, like the rest of the world, needs strong local neocloud providers. AI workloads are inherently private and sensitive, and enterprises and governments are increasingly reluctant to place critical data on foreign cloud platforms.
“This reluctance has created a strong push for sovereign, localised AI infrastructure. As a result, new platforms are being built from the ground up, combining deep infrastructure expertise with advanced cloud orchestration and AI-native design,” said Narendra Sen, founder and CEO of NeevCloud.
NeevCloud’s neocloud operations are headquartered in Bengaluru, with a 150-member team. Sen added that the company is building a full-stack AI cloud platform by expanding technology stack, and localising inference, orchestration, and GPU-as-a-service offerings.
To be sure, neocloud adoption has followed a phased pattern: startups and enterprises were early adopters, followed by academic institutions conducting advanced research and large-scale training. More recently, government programs—including initiatives under the IndiaAI Mission—have emerged as significant users, experts say.
“Yotta’s Shakti Cloud GPU-as-a-Service fits within this neocloud model and reflects how the category has evolved. Early neoclouds focused on bare-metal and GPU clusters to address access and performance gaps that hyperscalers were not designed for. Over time, the model has moved up the stack, combining GPU-first infrastructure with cloud-like usability, while avoiding rigid capacity structures and opaque pricing,” said Bhavesh Adhia, chief AI officer and chief strategy officer at Yotta Data Services.
Yotta has been empanelled under the IndiaAI Mission for offering GPU compute services at subsidised rates. According to its website, Shakti Cloud is used by listed giant Fractal Analytics, Sarvam AI, Qure AI, and government led projects like AI Bhashini, among others.
A Neocloud Future
Setting up a neocloud is a capital intensive business comprising massive upfront investments in AI-focused GPU infrastructure and high-speed networking.
As per JLL’s report, “Building GPU infrastructure requires substantial investment. Securing funding depends on demonstrating viable business models and key customer deals, often necessitating guidance from capital market consultants.” The economics become more fragile considering investment in data centers designed for high-density power, often 50-150kW per rack.
“While early movers may gain an initial edge, global cloud providers have far greater financial firepower, making consolidation inevitable. We are already seeing rising private equity funding and acquisition interest in India, driven by strong growth forecasts. Over time, we expect more takeovers, partnerships, and strategic investments as global players seek to expand their presence,” said Abhinav Johri, Partner at EY.
Typically, global providers prefer partnerships as the fastest route to market entry, a trend that has been repeatedly seen in India, he added. One of such examples is Microsoft’s major investments in CoreWeave and Nebius Group. An evolving ecosystem where hyperscalers and neocloud providers increasingly coexist and collaborate, can be expected.
The post India’s Neocloud Players Ride The Sovereign AI Wave appeared first on Inc42 Media.
https://ift.tt/KMPpAoS
0 Comments