The popular framing of the Nvidia (NVDA) debate — will artificial-intelligence (AI) demand hold up? — is the wrong question. For 2026 and 2027, demand is the settled part of the story: Nvidia is sitting on a roughly $500 billion Blackwell and Rubin order backlog and has guided to more than $1 trillion of demand visibility through 2027. The real variable in any honest Nvidia (NVDA) price prediction for 2026, 2028 and 2030 is not whether revenue grows, but what price-to-earnings multiple the market is willing to pay as growth decelerates from triple digits toward a still-large 20-25% a year. That single distinction explains why credible NVDA targets for 2030 range from under $300 to well above $900.
Here is the insight most NVDA forecasts bury: earnings and the multiple are now pulling in opposite directions. Nvidia’s profits are still compounding fast, but the market is unlikely to keep paying 35-40 times earnings forever for a company whose growth rate is normalising. Having tracked the AI-capex cycle since the first Hopper shortage, the cleanest historical parallel is Cisco Systems in 2000 — then the world’s most valuable company and the undisputed picks-and-shovels winner of the internet build-out. Cisco’s revenue kept rising for two decades, yet the stock did not reclaim its March 2000 peak for roughly 21 years, because its multiple collapsed. Nvidia is a far more profitable business than Cisco ever was, but the lesson stands: for an infrastructure leader trading at a premium multiple, the valuation, not the demand, is where the money is won or lost.
Key Facts:
• Nvidia (NVDA) traded at $192.53 with a market cap near $4.66 trillion on June 26, 2026 — StockAnalysis
• Fiscal Q1 2027 revenue was $81.6 billion, up 85% year over year, with data-center revenue of $75.2 billion — reported late May 2026
• The consensus 12-month target is roughly $300, in a $250-$500 range across 38 analysts rating NVDA a Strong Buy — MarketBeat
• Blackwell and Rubin order backlog is about $500 billion, with $1 trillion-plus demand visibility cited through 2027 — Wedbush, Citi, BofA
• FY2027 revenue consensus is about $391.3 billion, with EPS near $9.34 — Simply Wall St
• Roughly $50 billion in annual China data-center revenue is excluded from FY2027 guidance — Nvidia
• Third-party long-range forecasts span about $382-$612 (2028) and $678-$1,240 (2030) — Long Forecast
What’s actually happening and why
Nvidia enters the second half of 2026 as the most valuable public company in the world, with the share price near $192.53 and a market capitalisation of roughly $4.66 trillion as of June 26, 2026. The fundamentals remain extraordinary: fiscal first-quarter 2027 revenue, reported in late May 2026, came in at $81.6 billion, up 85% year over year, of which data-center revenue alone was $75.2 billion. For the full FY2026, revenue was $215.94 billion, a 65% increase on the prior year’s $130.50 billion.
The engine is the AI data-center build-out. Hyperscalers, sovereign-AI programmes and neoclouds are absorbing Blackwell-generation GPUs as fast as Taiwan Semiconductor Manufacturing Company (TSMC) can package them, and the next architecture, Vera Rubin, is already spoken for. The order book — about $500 billion across Blackwell and Rubin — is what gives analysts unusual confidence in the near term. It also reframes the bull case: this is less a bet on AI adoption than on Nvidia’s ability to convert a known backlog into shipped silicon at high margins, against the supply ceiling set by TSMC’s advanced packaging, a constraint we examine in our TSMC stock forecast.
That backdrop is why the Street remains overwhelmingly positive. As Dan Ives, managing director at Wedbush Securities, put it: “GTC 2026 was another opportunity for Jensen & Co. to further separate from the field in the AI arms race and they delivered, further reinforcing that Nvidia sits alone at the top of the AI mountain with the entire tech world watching below.” Ives, who carries a $275 target, has argued that Wall Street is “significantly underestimating” Nvidia’s demand drivers.
Wall Street and competitor response
The analyst community is close to unanimous on direction and split on magnitude. Across 38 analysts, NVDA holds a Strong Buy consensus with an average 12-month target around $300, a high of $500 and a low of $250. Morgan Stanley’s Joseph Moore frames the distribution most usefully: a $250 base case, an upside scenario toward $330 if Nvidia executes its roadmap, and a downside near $150 if AI infrastructure spending slows faster than expected. That $150-to-$330 spread on a 12-month view is the honest measure of the uncertainty — and it is driven almost entirely by the multiple, not by disagreement over near-term revenue.
The competitive response is the other half of the picture. Advanced Micro Devices (AMD) is pushing its MI-series accelerators as the credible second source, the case we lay out in our AMD stock forecast. Broadcom (AVGO) is arming the hyperscalers with custom AI silicon — application-specific integrated circuits (ASICs) that let Google, Meta and Amazon route their highest-volume inference workloads away from Nvidia GPUs, a structural threat detailed in our Broadcom stock forecast. Nvidia’s answer has been to sell systems, not chips — full NVLink racks, networking and the CUDA software moat — which keeps switching costs high even as raw silicon competition intensifies. None of this dents 2026; all of it matters for the multiple investors will pay in 2028 and 2030.
Nvidia (NVDA) price prediction: 2026, 2028 and 2030 scenarios
The table below is scenario analysis, not a guarantee. It applies a range of price-to-earnings multiples to consensus and extrapolated earnings-per-share (EPS) estimates, with EPS growing from the FY2027 consensus of about $9.34 at roughly 22-25% a year — the pace Simply Wall St models for Nvidia’s earnings. The base case assumes the multiple gradually compresses as growth normalises; the bull case assumes the market keeps paying a premium; the bear case is a Cisco-style multiple reset.
| Year | Bear case | Base case | Bull case | Base-case math |
|---|---|---|---|---|
| 2026 | $150 | $250 | $330 | ~$9.34 EPS × ~27x |
| 2028 | $230 | $380 | $560 | ~$14 EPS × ~27x |
| 2030 | $300 | $520 | $900 | ~$21 EPS × ~25x |
The synthesis those numbers produce is the heart of this Nvidia (NVDA) price prediction. In the base case, EPS more than doubles between 2026 and 2030, yet the share price only roughly doubles — because the multiple compresses from about 27 times to about 25 times and the starting point already embeds optimism. The bull case requires the market to keep paying 35-42 times earnings into the end of the decade, which is plausible only if AI demand stays supply-constrained. The bear case does not require a demand collapse at all; it merely requires the multiple to fall toward 14-16 times, exactly what happened to Cisco even as its business kept growing. That is the information-gain point competitors miss: you can be right about Nvidia’s earnings and still be wrong about the stock.
There is a subtler data point that cuts against the bears, however. At $192.53 against the FY2027 consensus EPS of about $9.34, Nvidia already trades at roughly 21 times forward earnings — below the 27 times used in the base case and well beneath the 35-40 times it commanded through 2024 and 2025. In other words, the market has already done part of the multiple compression the bear case fears. Combine that with gross margins still around 75% and a backlog that de-risks the next two years of revenue, and the synthesis is nuanced: NVDA is no longer priced as a runaway momentum stock, but as a high-quality compounder whose forward multiple is closer to the broad market than the headlines suggest. The bull-bear tension is therefore narrower than the raw price range implies — most of the disagreement is about 2028 onward, when the order book stops providing cover.
The two cases can be set side by side:
| Bull case for NVDA | Bear case for NVDA |
|---|---|
| ~$500bn Blackwell/Rubin backlog de-risks 2026-2027 revenue | Growth decelerates from 85% toward the mid-20s%, inviting a re-rating |
| CUDA software moat and full-rack systems keep switching costs high | Custom ASICs (Broadcom) and AMD divert high-volume inference workloads |
| Forward P/E already near 21x — premium has partly deflated | Cisco precedent: revenue rose for years while the stock stalled on multiple compression |
| H200 China restart is upside not in consensus | ~$50bn China revenue excluded; tighter export controls widen the gap |
The China and regulatory tension
The single largest swing factor outside the multiple is policy. US export controls administered by the Commerce Department’s Bureau of Industry and Security have already carved roughly $50 billion of annual China data-center revenue out of Nvidia’s FY2027 guidance, and the political pressure is rising. Senator Elizabeth Warren invited Jensen Huang to a Senate hearing on China AI-chip sales; Huang declined to testify on June 8, 2026. The tension is genuine: Washington wants to slow China’s AI compute, while Nvidia wants to defend a market it is losing to Huawei.
Huang has been blunt about the cost. “The demand in China is quite large,” he told CNBC, adding that “Huawei is very, very strong” and that its “local ecosystem of chip companies are doing quite well, because we’ve evacuated that market.” (CNBC) The read-through for a multi-year forecast is asymmetric: the bear case must assume controls tighten and the China gap widens, while a partial thaw — Nvidia has restarted H200 manufacturing for approved Chinese orders — would be upside not currently in consensus numbers. For brokers and institutional desks modelling NVDA, China is a binary overlay on an otherwise supply-driven story.
What happens next — predictions
Three calls follow from the framework. First, through 2026 the stock tracks the backlog and the macro: with the order book locked, NVDA’s path is set more by the rate environment and risk appetite than by demand surprises, which is why the base case sits near the $250 consensus rather than the $300-plus high targets. Second, 2028 is the multiple-compression year — as year-over-year growth slows from 80%-plus toward the mid-20s, expect the market to re-rate the P/E lower even as EPS climbs, capping the base case near $380 despite far higher profits. Third, by 2030 the dispersion is enormous because it is almost entirely a valuation question: the same roughly $21 of EPS supports $300 in a Cisco-style reset and $900 if Nvidia is still treated as a premium compounder.
For longer-horizon investors, the discipline is to separate the company from the stock. Nvidia the business is very likely larger, more profitable and still dominant in 2030. Nvidia the stock depends on a multiple no model can pin down — which is exactly why a credible forecast must be a range, not a point. The comparable AI-infrastructure names round out the picture in our Intel stock forecast.
FAQ
What is the Nvidia (NVDA) price prediction for 2026?
This analysis models a 2026 base case near $250, a bull case around $330 and a bear case near $150, anchored to Morgan Stanley’s scenarios and the $250-$500 Street range. The consensus 12-month target across 38 analysts is roughly $300, with the stock at $192.53 on June 26, 2026.
Where could NVDA stock be in 2030?
The 2030 base case here is about $520, with a bull case near $900 and a bear case around $300. The wide range reflects valuation, not demand: the same roughly $21 of projected EPS supports very different prices depending on the price-to-earnings multiple the market pays.
Why is the multiple more important than AI demand?
Near-term demand is largely locked by a $500 billion Blackwell and Rubin backlog. The bigger uncertainty is whether the market keeps paying 35-40 times earnings as growth slows. History — notably Cisco after 2000 — shows an infrastructure leader’s revenue can keep rising while the stock stalls on multiple compression.
How does China affect the Nvidia forecast?
US export controls have excluded roughly $50 billion of annual China data-center revenue from FY2027 guidance. Tighter controls widen the bear case, while an approved restart of H200 sales to China — already under way — is potential upside not fully reflected in consensus estimates.
Is Nvidia stock a buy at these levels?
That depends on your view of the multiple and your risk tolerance, not on this article. The bull-bear spread is unusually wide precisely because reasonable investors disagree on valuation. This piece is informational analysis, not a recommendation; see the disclaimer below.
What are the biggest risks to this Nvidia forecast?
The primary risk is valuation: a fall in the price-to-earnings multiple toward the mid-teens would cap or cut the stock even if revenue keeps rising, the Cisco-after-2000 scenario. Secondary risks include a faster-than-expected slowdown in AI capital spending by hyperscalers, share loss to custom ASICs and AMD, tighter US-China export controls widening the roughly $50 billion China gap, and any slip in TSMC’s advanced-packaging capacity that constrains Blackwell and Rubin shipments.
Disclaimer: This article is informational analysis only and is not financial, investment, or trading advice. Equities are volatile and can lose value rapidly; price targets are scenario estimates based on stated assumptions and are not guarantees. Past performance and analyst forecasts do not assure future results. Do your own research and consult a regulated financial adviser before making any investment decision.