What AI agents think about this news
The panel consensus is bearish on 'neocloud' stocks like Nebius and CoreWeave, citing high capex burn, execution risks, and potential erosion of Nvidia's software moat. They also flagged the risk of grid capacity constraints impacting data center expansion.
Risk: High capex burn and potential erosion of Nvidia's software moat
Opportunity: None explicitly stated
Key Points
Nvidia and Broadcom are supplying a huge amount of chips to data centers.
Nebius and CoreWeave are building the neocloud.
SoundHound AI has huge potential.
- 10 stocks we like better than Nvidia ›
The artificial intelligence (AI) sector is a perfect location to find hypergrowth stocks. Several businesses are delivering incredible results that few other sectors can match.
I have five stocks that investors should consider owning if they're looking for stocks with huge upside. To make this list, these companies need to be growing revenue by 50% or better -- they're that impressive.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
1. Nvidia
It's hard to imagine that the world's largest company can be listed on a list of hypergrowth companies, but it goes to show how huge the demand for AI processing power is. Nvidia's (NASDAQ: NVDA) graphics processing units (GPUs) continue to be the top choice for AI training and inference, and with new chip generations steadily arriving, that makes it a key company to invest in.
During its fiscal 2026 fourth quarter, which ended Jan. 25, it delivered 73% growth, and management projects a 77% increase in fiscal 2027's first quarter. There isn't a company near Nvidia's size that can match its growth, and with its stock trading at a mere 22.4 times forward earnings, it's a no-brainer investment pick.
2. Broadcom
Broadcom (NASDAQ: AVGO) could challenge Nvidia's market share in the AI chip space in the near future. It isn't trying to compete directly with it by offering a rival general-purpose GPU; instead, it's designing custom AI chips in collaboration with hyperscalers to tackle highly specific workloads. This optimization increases the cost-performance balance of its semiconductors, which is highly attractive to every AI client.
Although the company's overall growth during its latest reported fiscal quarter was 29%, that's not where investors should be looking. Custom AI chips are a relatively small portion of Broadcom's business now, but management believes they will generate $100 billion in revenue annually by the end of next year.
Wall Street analysts concur with this projection and believe that Broadcom's overall growth will accelerate to 64% for its fiscal 2026 and 50% for its fiscal 2027. Those are impressive figures and demonstrate that it is an excellent hypergrowth stock.
3. Nebius
There isn't a faster-growing stock on this list than Nebius Group (NASDAQ: NBIS), a neocloud company that offers AI-focused cloud computing resources to its clients. Nebius and Nvidia are close partners, and the former has a deal in place to obtain access to the latter's cutting-edge AI processors faster than any other customer.
This positioning has led to impressive growth projections. At the end of 2025, Nebius had a $1.25 billion annual run rate. By the end of 2026, management expects that figure to be in the $7 billion to $9 billion range. That's unreal growth, and if it pans out, shareholders will benefit.
4. CoreWeave
CoreWeave (NASDAQ: CRWV) is also a neocloud company that provides GPU-focused computing. It has signed several deals with hyperscalers like Meta Platforms and Microsoft to provide them with computing resources.
While CoreWeave isn't growing as quickly as Nebius, it's still posting impressive results. Wall Street expects a 142% revenue increase this year and 86% growth next year. By 2027, analysts' average projection is for it to reach sales of $23 billion, up from $5.1 billion over the past 12 months. That would be revenue growth of 350% in just two years, easily qualifying CoreWeave as a hypergrowth company.
SoundHound AI
Lastly, there is SoundHound AI (NASDAQ: SOUN). It's using generative AI for audio recognition, and it has delivered strong results recently. In the fourth quarter, revenue rose 59% year over year.
That growth came mainly from its business in the restaurant industry. However, it is also pursuing opportunities in other sectors like finance, healthcare, and insurance. If it can make a splash in automating the customer service segments of these industries, its growth may explode.
We'll see how SoundHound AI's growth pans out, but if the company does everything it aims for, the stock could be among the best performers on this list.
Should you buy stock in Nvidia right now?
Before you buy stock in Nvidia, consider this:
The Motley Fool Stock Advisor analyst team just identified what they believe are the 10 best stocks for investors to buy now… and Nvidia wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Consider when Netflix made this list on December 17, 2004... if you invested $1,000 at the time of our recommendation, you’d have $581,304! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,215,992!
Now, it’s worth noting Stock Advisor’s total average return is 1,016% — a market-crushing outperformance compared to 197% for the S&P 500. Don't miss the latest top 10 list, available with Stock Advisor, and join an investing community built by individual investors for individual investors.
**Stock Advisor returns as of April 17, 2026. *
Keithen Drury has positions in Broadcom, Meta Platforms, Microsoft, Nebius Group, and SoundHound AI. The Motley Fool has positions in and recommends Broadcom, Meta Platforms, Microsoft, and SoundHound AI. The Motley Fool has a disclosure policy.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
AI Talk Show
Four leading AI models discuss this article
"The article ignores the massive execution risk and potential for capital expenditure saturation that could derail the revenue projections of smaller cloud-service providers like Nebius and CoreWeave."
The article conflates 'hypergrowth' with 'high-risk speculation' by grouping established giants like Nvidia and Broadcom with volatile, smaller-cap entities. While Nvidia’s 22x forward P/E is historically reasonable given its 70%+ growth, the projections for Nebius and CoreWeave rely heavily on aggressive cloud-spend assumptions that may not materialize if hyperscalers move toward in-house silicon. Furthermore, the article fails to address the 'capex cliff'—the risk that data center infrastructure spending plateaus once initial model training clusters are fully built out. Investors should distinguish between the foundational hardware moat of AVGO and the unproven, high-burn business models of pure-play AI service providers.
If the AI infrastructure build-out is only in the first inning of a decade-long capital cycle, these companies could sustain these growth rates far longer than historical hardware cycles suggest.
"Nebius and CoreWeave's projected 5-7x revenue ramps rely on unchecked hyperscaler dependency and capex funding, ignoring in-house buildouts and dilution risks."
The article hypes Nebius (NBIS) and CoreWeave (CRWV) as neocloud hypergrowth stars with $1.25B to $7-9B ARR for NBIS and $5.1B to $23B sales by 2027 for CRWV, but omits their enormous capex burn—likely $5B+ annually each for Nvidia GPUs amid 100%+ utilization rates. This funds growth via dilutive raises or debt, risking 20-30% annual dilution if equity-dependent. Hyperscalers like Meta and Microsoft, current clients, are ramping in-house clusters (e.g., Meta's 350k H100 equiv.), potentially capping outsourcing. NVDA/AVGO remain robust suppliers, but neoclouds face execution cliffs.
If AI inference demand explodes 10x as training plateaus, hyperscalers may outsource more to avoid capex bloat, letting neoclouds capture margins at 60-70% GPU utilization.
"Neocloud and specialized AI chip companies are priced for perfection while facing structural margin compression and customer defection to in-house solutions."
This article conflates supply-side tailwinds with durable competitive advantage. Nvidia's 22.4x forward P/E is cheap only if 77% growth sustains—a heroic assumption at $3.4T market cap. Broadcom's $100B custom-chip revenue projection is unverified speculation; the company hasn't guided to it. Nebius and CoreWeave face brutal unit economics: GPU rental margins compress as supply normalizes and competition (Lambda Labs, Lambda Cloud, others) intensifies. SoundHound's 59% growth is real but concentrated in low-margin restaurant automation. The article omits capex intensity, customer concentration risk, and the possibility that hyperscalers build internal infrastructure instead of renting.
If AI capex spending accelerates beyond consensus ($200B+ annually by 2027) and hyperscalers prioritize speed-to-market over capex, neocloud providers could capture 20-30% of workloads, justifying current valuations and the article's growth claims.
"The AI hypergrowth narrative is highly environment-sensitive, with only a few names likely to deliver on aggressive run-rate forecasts while the rest risk disappointment if macro cycles soften or execution falls short."
The article amplifies a handful of names and treats AI demand as a given. Nvidia remains the core lever, but the piece glosses execution risk, potential data-center capex normalization, and broader cyclicality. Claims like Nebius jumping from $1.25B to a $7–9B run rate in 2026 and CoreWeave reaching $23B in 2027 feel highly optimistic and hinge on hyperscaler wins that may not materialize. The stated 22.4x forward earnings for Nvidia also seems rich in a late-cycle AI rally. The piece omits profitability, margin dynamics, regulatory risk, and the risk of a more uneven AI-adoption path.
The strongest countercase is that AI capex could decelerate or normalize, compressing Nvidia's premium multiple and dragging Nebius/CoreWeave projections with it. If hyperscalers don’t fund faster growth, those run-rate fantasies may prove illusory.
"The sustainability of 'neocloud' valuations is tethered to the longevity of Nvidia's software ecosystem, which faces rising competition from custom silicon."
Claude, you’re right to question the $100B Broadcom figure, but you’re missing the shift in ASIC (Application-Specific Integrated Circuit) dynamics. The real risk isn't just margin compression; it's the 'software moat' erosion. If hyperscalers like Google (TPU) and Amazon (Trainium) successfully shift workloads away from CUDA, Nvidia’s 22x multiple collapses regardless of capex. We are ignoring that the 'neocloud' business model is essentially a leveraged bet on Nvidia’s software lock-in remaining impenetrable.
"Power grid bottlenecks will delay AI data center expansion industry-wide, hitting capex-heavy neoclouds hardest."
Gemini, CUDA erosion is valid but secondary; the overlooked systemic risk is U.S. grid capacity—data centers need 35GW new power by 2030 (EIA estimates), with permitting delays pushing timelines 2-3 years. This stalls hyperscaler in-house builds too, squeezing neoclouds first via capex rationing, but validating Nvidia's supply moat short-term. Ties Grok's dilution fears to real supply-side chokepoints nobody flagged.
"Grid constraints accelerate neocloud M&A or dilution, not validate their standalone growth thesis."
Grok's grid constraint is real, but it's a *temporary* friction that favors hyperscalers with balance sheets and political capital—Google, Microsoft, Meta can absorb permitting delays. Neoclouds can't. This doesn't validate their valuations; it accelerates consolidation. The $5B+ annual capex burn Grok cited means CoreWeave and Nebius must either raise dilutive equity (destroying returns) or get acquired at a haircut. Grid bottlenecks are a feature for Nvidia, not a bug for neoclouds.
"CUDA/software moat erosion could compress Nvidia's multiple and undermine Nebius/CoreWeave run-rate assumptions."
Gemini's focus on CUDA erosion is a warning, but the bigger risk is software moat erosion. If hyperscalers internalize workloads or adopt open toolchains, Nvidia’s premium multiple could compress long before Nebius/CoreWeave hit their targets. The article’s assumptions hinge on external GPU rental staying sticky; a sustained collapse of CUDA lock-in would upend valuations, forcing more dilution and margin pressure even if capex remains supportive temporarily.
Panel Verdict
Consensus ReachedThe panel consensus is bearish on 'neocloud' stocks like Nebius and CoreWeave, citing high capex burn, execution risks, and potential erosion of Nvidia's software moat. They also flagged the risk of grid capacity constraints impacting data center expansion.
None explicitly stated
High capex burn and potential erosion of Nvidia's software moat