Investors Eye “Neoclouds” as AI Cloud Specialists Amid Emerging Risks
“Neoclouds” are gaining strong investor interest as specialized AI cloud computing providers. Unlike traditional hyperscalers such as Amazon and Microsoft, neoclouds focus exclusively on AI workloads, handling hardware and software installations with greater speed, customization, and potentially better cost and energy efficiency for clients.
CoreWeave, which went public in March, stands as the largest listed neocloud. Nebius is another rising contender, while private companies like Lambda and Crusoe attract venture capital and secure major AI contracts. Nvidia has also deepened its involvement by investing in these startups and ramping up its own cloud services amid growing competition from hyperscalers.
Balancing Growth and Financial Risks
Analysts appreciate the neocloud model’s potential to capture AI cloud market share but caution about looming challenges. High capital expenditure (capex), limited differentiation, and inflated AI valuations pose significant risks. According to UBS strategist Sundeep Gantori, heavy capex pressures margins despite the promise of future returns.
UBS projects AI-related capex outside the big four tech giants—Amazon, Microsoft, Meta, and Alphabet—to rise 85% year-over-year, reaching $150 billion by 2025. This figure includes spending from China, neoclouds, other hyperscalers, and various enterprise and sovereign cloud providers.
Market Outlook and Long-Term Potential
Goldman Sachs offers a broader perspective, estimating the GPU-as-a-Service and AI cloud market could reach $267 billion by 2030. They anticipate neoclouds will command about 20% of this market, while hyperscalers retain the remaining 80%.
The Nvidia-Neocloud Relationship: Opportunity and Risks
Nvidia’s strategic investments in CoreWeave, Lambda, Crusoe Energy, and Nebius have accelerated neocloud growth and strengthened Nvidia’s competitive position. HSBC analyst Abhishek Shukla highlights that while this strategy broadens Nvidia’s GPU market, it also squeezes neocloud profit margins as GPU leasing prices drop.
Shukla warns that neoclouds face the danger of losing key customers if hyperscalers begin sourcing GPUs directly from Nvidia. For example, CoreWeave’s business with Microsoft may depend on current GPU shortages and Nvidia’s preferential GPU allocations. Should these factors change, Microsoft and OpenAI might bypass neoclouds entirely, dealing directly with Nvidia.
Additionally, over half of CoreWeave’s debt financing comes from original equipment manufacturers (OEMs), including Nvidia. This setup benefits Nvidia only if CoreWeave continues to lease GPUs competitively, fostering a more dynamic GPU leasing market.
Neoclouds and Nvidia’s Growth Strategy
Some analysts see Nvidia’s support of neoclouds and its own DGX AI platform as a way for the company to monetize large-scale AI inference workloads beyond GPU sales. Loop Capital analyst Ananda Baruah notes that Nvidia now treats neoclouds as an extension of its data center strategy, aligning with its inferencing ambitions.
Wall Street’s Take on Leading Neoclouds
CoreWeave and Nebius stand out as the main publicly recognized neoclouds. Nebius, based in Europe, currently enjoys more favorable analyst coverage. Goldman Sachs analyst Alexander Duval initiated coverage on Nebius with a buy rating and a 12-month price target implying 25% upside, citing its full-stack software, cost advantages, and growing AI customer base.
CoreWeave, valued at over $62 billion, has seen its shares surge more than 200% since its IPO but dropped 30% in the last month. The decline followed CoreWeave’s announcement to acquire Core Scientific in a roughly $9 billion all-stock deal, raising concerns about valuation and stock dilution.
HSBC’s Shukla initiated CoreWeave coverage with a reduce rating and a $32 price target, describing the stock as significantly overvalued. He points out that CoreWeave’s revenue relies heavily on Microsoft and OpenAI, companies that mostly use their own software, weakening CoreWeave’s unique value. Liquidity issues and high borrowing costs add to the challenges.
Other firms like Mizuho Securities and Stifel have downgraded CoreWeave to hold, while Citigroup flagged downside risks. CFRA analyst Angelo Zino maintains a cautious stance, acknowledging CoreWeave’s revenue backlog and potential operating profit by 2026 but concerned about valuation and prefers alternative investment options in AI compute.
Private Neocloud Innovators and Their Growing Roles
The neocloud sector also includes private startups such as Lambda, Covalent, and Crusoe Energy. Crusoe, which repurposes flare gas to power data centers, was selected by Oracle to build a major campus for OpenAI’s Stargate project in Texas. This campus is expected to house up to 400,000 Nvidia Blackwell GPUs. Like CoreWeave, Crusoe pivoted from crypto mining to AI-focused infrastructure.
Lambda offers competitively priced Nvidia GPUs alongside cloud services and software to train and deploy AI models, expanding options for AI enterprises seeking specialized compute capacity.
Conclusion
Neoclouds present an intriguing investment opportunity amid the AI cloud surge, offering specialized services that differ from traditional hyperscalers. However, investors should weigh the risks of heavy capex, valuation uncertainty, and Nvidia’s dominant role in the GPU supply chain. Monitoring how hyperscalers adjust GPU sourcing strategies will be key to assessing neoclouds’ long-term viability.
For those interested in AI cloud computing and infrastructure, exploring comprehensive AI training courses can provide valuable insights into the technology and market dynamics. Check out Complete AI Training’s latest courses for practical knowledge on AI development and deployment.
Your membership also unlocks:
