By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
Despite Nvidia's impressive growth and seemingly attractive valuation, the panel consensus is bearish due to significant risks, including the 'law of large numbers', extreme customer concentration, geopolitical export controls, and the threat of custom ASICs eroding GPU demand and margins.
Risk: Structural shift toward internal silicon at major hyperscalers, potentially leading to a 'margin-squeeze' scenario and evaporating Nvidia's pricing power.
Opportunity: None explicitly stated; panel focused on risks and challenges.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
Nvidia is a dominant force in semiconductors.
It's growing in new directions now.
Its stock is appealingly valued, given its growth rate.
Check out how magnificently Nvidia (NASDAQ: NVDA) stock has performed in recent years:
| Time period | Average annual return | |---|---| | Past 1 year | 62.22% | | Past 3 years | 78.30% | | Past 5 years | 67.76% | | Past 10 years | 68.93% | | Past 15 years | 50.68% |
That's enough to turn a single $10,000 investment into more than $130,000 over the past five years, or roughly $1.9 million over the past decade.
That might have you wondering if you missed this boat. Surprisingly, the shares still seem attractively valued.
Check it out: Nvidia's recent forward-looking price-to-earnings (P/E) ratio of 25.6 is well below its five-year average of 36.1. And its recent price-to-sales ratio of 21.5 is below its five-year average of 24. (Those price-to-sales numbers are steep!) Steep valuation measures can be justified, though, when a company is growing briskly.
In its recently reported first quarter, Nvidia posted revenue of $81.6 billion, up 85% from the year-ago quarter. The company has found great success supplying chips to gobs of data centers that are facilitating artificial intelligence (AI) processing, and its first quarter also featured record data center revenue of $75.2 billion, up 92% year over year.
The future matters more than the past to investors, though, and on this count, too, Nvidia looks good -- in part because it's expanding in new directions, such as entering the central processing unit (CPU) market with a new chip. Its data center business is poised to keep profiting handsomely as well, as major tech companies are spending hundreds of billions on data center capital expenditures and are projected to spend trillions on it within a few years.
Given Nvidia's market dominance, promising future, and attractive valuation, I think that it still has plenty of room to run. (The stock and/or the market could pull back in the short term, which is why I suggest maintaining a long-term focus.)
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 $449,393! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,366,006!
Now, it’s worth noting Stock Advisor’s total average return is 983% — a market-crushing outperformance compared to 212% 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 June 4, 2026. *
Selena Maranjian has positions in Nvidia. The Motley Fool has positions in and recommends Nvidia. 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.
Four leading AI models discuss this article
"Nvidia's upside relies on an uninterrupted AI data-center capex boom; any slowdown or policy barrier could crush the valuation even if earnings remain solid."
The article frames Nvidia as a dominant AI hardware winner with a durable data-center ramp and a currently reasonable forward multiple. It cites Q1 revenue of $81.6B (up 85% YoY) and data-center revenue of $75.2B (up 92%), plus a lower forward P/E of 25.6 versus a 5-year average of 36.1. Yet the bullish case hinges on two fragile bets: ongoing, unbroken AI data-center capex and Nvidia’s ability to sustain growth after CPU entry and other expansions, without implying margin pressure or intensifying competition. The piece also glosses regulatory/regional risks (e.g., export controls) and potential cyclicality in AI spend, which could trigger multiple compression even with solid earnings.
The bull case assumes perpetual AI demand and a never-ending data-center capex spree; a slowdown in enterprise AI budgets or regulatory headwinds (like China export controls) could prompt a sharp re-rating despite earnings staying resilient.
"Nvidia's valuation is currently priced for perfection, failing to account for the inevitable deceleration of hyper-growth as capital expenditure cycles normalize."
Nvidia's current forward P/E of ~25x appears deceptively cheap only if you accept the sell-side's aggressive consensus estimates for fiscal 2026. The article ignores the 'law of large numbers'; sustaining 85% revenue growth on a $80B+ base is mathematically improbable as hyperscalers like Microsoft and Meta inevitably optimize their AI spend. Furthermore, the reliance on a few concentrated customers for the majority of data center revenue creates significant tail risk. While the move into CPUs is a strategic pivot, it pits them against entrenched incumbents like Intel and AMD, likely compressing margins. I view the current valuation as a 'peak earnings' trap rather than a genuine discount.
If AI infrastructure becomes the new 'electricity' of the global economy, Nvidia’s current valuation is merely the entry price for a multi-decade monopoly on the fundamental compute layer of the future.
"NVDA's valuation is cheap relative to its recent history, not relative to deceleration risk and the cyclical nature of capex spending."
The article conflates valuation cheapness with safety. Yes, NVDA's 25.6x forward P/E is below its 5-year average of 36.1—but that average was inflated by speculative excess, not justified by fundamentals. The real issue: 85% YoY revenue growth and 92% data center growth are decelerating from 2023 peaks. If Q2 guidance disappoints or growth slows to 40-50%, the stock reprices sharply lower despite 'attractive' multiples. The CPU ambitions are unproven against entrenched AMD/Intel. The article also ignores capex cycles—when hyperscalers pause spending (cyclical risk), NVDA's moat narrows temporarily. Finally, $1.9M on $10k over a decade is backward-looking; past performance doesn't predict forward returns.
NVDA's dominance in AI inference/training chips, combined with trillions in projected capex, genuinely could justify 25-30x multiples for another 3-5 years if execution holds and market share expands.
"Nvidia's valuation embeds assumptions of sustained 80%+ growth that face material risks from capex pauses and custom-chip competition."
Nvidia's 85% revenue jump to $81.6B and 25.6x forward P/E appear compelling, yet the piece ignores extreme concentration: data centers delivered 92% of growth and $75.2B. Hyperscaler capex plans can be deferred or redirected to custom ASICs from Microsoft, Google, and Amazon, eroding GPU demand. Geopolitical export controls have already disrupted China sales repeatedly, and semiconductor cycles show 50%+ growth rarely persists beyond a few years. A normalization to 25-30% growth would justify a lower multiple than the current 21.5x sales.
Nvidia's CUDA software lock-in could still preserve 70%+ gross margins and pricing power even if unit growth slows, allowing the stock to compound if AI adoption exceeds current forecasts.
"Overlooked risk: price/mix and ecosystem lock-in could erode margins if hyperscalers pivot to in-house accelerators, pressuring ASP and growth even with strong usage."
Gemini’s 'law of large numbers' critique is fair, but the overlooked risk is price/mix and ecosystem lock-in. If hyperscalers push in-house accelerators or alternative architectures, Nvidia could face ASP pressure and slower unit growth even with solid usage. Add export-regulatory tailwinds/cycles and a modest 2H24 slowdown could trigger outsized multiple compression. The bullish case rests on execution and moat expansion, not just demand.
"The rise of internal custom ASICs at major hyperscalers poses a greater threat to Nvidia's margins than cyclical spending fluctuations."
Grok, your focus on custom ASICs is the missing link. While Claude worries about cyclicality, the real threat to Nvidia’s 70%+ gross margins isn't just a spending pause—it's the structural shift toward internal silicon at Google (TPU), Amazon (Trainium), and Microsoft (Maia). If these hyperscalers successfully transition even 20% of their inference workloads to custom chips, Nvidia’s pricing power evaporates. We are looking at a potential 'margin-squeeze' scenario, not just a unit growth slowdown.
"Custom ASICs threaten unit growth and inference revenue, but training workload stickiness buys Nvidia 18-24 months of pricing power—the real inflection point is whether hyperscalers crack training efficiency by late 2025."
Gemini's ASIC threat is real, but the timeline matters enormously. Google's TPU and Amazon's Trainium handle *inference*—lower-margin, lower-urgency workloads. Training still requires Nvidia's H100/H200 for 18-24 months minimum due to CUDA lock-in and software maturity. A 20% inference shift hurts, but doesn't crater margins if training ASPs hold. The bigger risk: if hyperscalers *successfully* train on custom silicon within 24 months, then we have a structural problem. Until then, Nvidia's gross margin floor is ~70%, not collapsing.
"Training migration to custom silicon could occur faster than 18-24 months, hitting margins sooner than Claude allows."
Claude's 18-24 month training lock-in timeline overlooks how quickly hyperscalers iterate once they allocate serious engineering resources. Microsoft's Maia and Amazon's Trainium already target mixed workloads; a successful 10-15% training shift by late 2025 would compress ASPs and volumes simultaneously, amplifying Gemini's structural margin risk beyond any cyclical pause or export-control noise.
Despite Nvidia's impressive growth and seemingly attractive valuation, the panel consensus is bearish due to significant risks, including the 'law of large numbers', extreme customer concentration, geopolitical export controls, and the threat of custom ASICs eroding GPU demand and margins.
None explicitly stated; panel focused on risks and challenges.
Structural shift toward internal silicon at major hyperscalers, potentially leading to a 'margin-squeeze' scenario and evaporating Nvidia's pricing power.