来自 Maksym Misichenko · Nasdaq ·
AI智能体对这条新闻的看法
The panel has mixed views on NVDA's $400 price target, with concerns about competition, potential capex slowdown, and the sustainability of high gross margins. While some panelists acknowledge NVDA's strong growth narrative and AI leadership, others warn of potential downside risks and the need for near-flawless execution.
风险: Rapid erosion of Nvidia's AI accelerator share due to competition from AMD, Intel, and custom ASICs, potentially leading to a cliff risk in earnings and multiple compression.
机会: Nvidia's CUDA software lock-in could delay share loss by 2-3 years, potentially allowing earnings to beat estimates and temporarily support the 44x multiple.
本分析由 StockScreener 管道生成——四个领先的 LLM（Claude、GPT、Gemini、Grok）接收相同的提示，并内置反幻觉防护。 阅读方法论 →
英伟达令人印象深刻的盈利增长表明该股票目前被低估。
该公司专注于扩展到新的人工智能利基市场，将使其比更广泛的市场以显着更快的速度增长。
英伟达 (纳斯达克：NVDA) 的股票在过去一年中上涨了 58%。 虽然这是一个令人印象深刻的增长，但与同期 PHLX 半导体板块 指数上涨 164% 相比，该股票的回报有所减弱。
我认为市场对英伟达股票未能充分奖励其季度季度交付的卓越增长。该公司最新的季度报告也没有改变其命运，尽管它在 5 月 20 日发布了 2027 财年第一季度业绩（截至 4 月 26 日）时，股价在超出华尔街预期后有所下跌。
人工智能会创造世界上第一个万亿美元富豪吗？ 我们的团队刚刚发布了一份报告，内容是关于一种鲜为人知的公司，被称为“不可或缺的垄断”，它提供英伟达和英特尔都需要的关键技术。继续 »
然而，不久之后英伟达将再次加速。事实上，这种半导体股票在一年内跳升至 400 美元的机会很大。让我们看看为什么可能就是这种情况。
英伟达是一家成长型股票，这意味着其收入和盈利增长速度远快于更广泛的市场。这从该公司的最新季度报告中可见一斑。英伟达 2027 财年第一季度（截至 4 月 26 日）的收入增长了 85%，达到创纪录的 816 亿美元。
其每股收益增长了 140%，达到 1.87 美元。这远高于根据 JPMorgan 预计今年第一季度标普 500 指数公司将实现的 13% 的盈利增长。更重要的是，英伟达的盈利额超过了第一季度科技行业估计的 45% 的增长。
重要的是，英伟达在过去一年中优于标普 500 指数的盈利增长。
对于英伟达投资者来说，好消息是其红热的盈利增长不会很快放缓。该公司对本季度 910 亿美元的收入指导预示着线上增长将同比增长 95%。此外，英伟达预计本季度非 GAAP 毛利率为 75%，比去年提高了 2.3 个百分点。因此，很容易理解为什么分析师预计英伟达的盈利将在本季度翻倍。
英伟达的高增长是由其在蓬勃发展的人工智能 (AI) 芯片市场中的巨大市场份额所驱动的。据报道，英伟达控制着人工智能加速器市场的 80% 到 90%，其加速增长率表明它并没有失去对这一有利可图的空间的控制。
这是因为英伟达正在针对人工智能中出现的新增长机会。例如，人工智能服务器中节能中央处理器 (CPU) 的需求解释了该公司现在将其服务器处理器作为独立产品提供。该公司最新一代 Vera 服务器 CPU 在客户中广受欢迎，英伟达预计今年将从该产品中获得 200 亿美元的收入。
这个数字表明英伟达可以多快地进入一个新领域并扩大规模。到目前为止，AMD 和英特尔一直主导着服务器 CPU 市场。AMD 预计到 2025 年，服务器 CPU 市场价值为 260 亿美元。因此，英伟达有望占据这一快速增长领域的很大一部分，预计这将为其长期开辟 2000 亿美元的收入机会。
此外，英伟达专注于通过进入新兴但有前景的物理人工智能市场来创造新的增长机会。该公司在过去四个季度中已经从物理人工智能市场中获得了超过 90 亿美元的收入。相比之下，英伟达在 2026 财年的物理人工智能收入为 60 亿美元，表明该业务的收入增长率上季度增长了惊人的 50%。
物理人工智能是指将人工智能与机器人、无人机和汽车等物体集成，使它们能够在现实世界中做出决策并执行复杂的操作。这个利基市场预计在长期内会稳步增长，因为它有可能提高工厂生产力、协助医疗专业人员并使用无人机交付产品。
根据沙特阿拉伯政府支持的风险投资倡议研究所 (Future Investment Initiative Institute)，物理人工智能市场在未来十年内可能会产生 2000 亿美元的收入，从目前的 20 亿至 30 亿美元大幅增长。因此，英伟达不仅专注于其已经服务的市场，而且还在努力定义应该允许其在长期内维持卓越增长的理想新类别。
根据涵盖该股票的 70 名分析师，英伟达的 12 个月中位数目标价为 293 美元。这表明从当前水平来看，潜在上涨空间为 37%。然而，投资者不应忽视分析师预计英伟达今年的盈利将增长 87%，达到每股 8.94 美元，几乎是根据 Yahoo! Finance 预计标普 500 公司将实现的 22% 的增长的 4 倍。
该人工智能股票以 24.3 倍的远期市盈率交易。这只是略高于标普 500 指数平均远期市盈率 21.8 倍的微小溢价。鉴于其盈利增长将超过标普 500 指数的平均水平 4 倍，英伟达应该获得优质估值。假设英伟达在一年后以 44 倍的市盈率交易，并且其底线达到每股 8.94 美元，其股价可能会跳升至 393 美元。
这在未来一年内潜在的上涨空间为 83%。然而，不要惊讶于英伟达今年由于其进入新市场以及其核心数据中心业务中令人印象深刻的积压，实现更强劲的盈利增长，这将足以将这家人工智能先驱的股价推升至 400 美元以上。
在您购买英伟达股票之前，请考虑以下几点：
Motley Fool Stock Advisor 分析师团队刚刚确定他们认为投资者现在应该购买的 10 支最佳股票……而英伟达不是其中之一。这些股票可能会在未来几年内产生巨大的回报。
请考虑 Netflix 在 2004 年 12 月 17 日被列入此名单时的情况……如果您当时投资了 1,000 美元，您将拥有 463,900 美元！或者当 英伟达 在 2005 年 4 月 15 日被列入此名单时……如果您当时投资了 1,000 美元，您将拥有 1,294,401 美元！
值得注意的是，Stock Advisor 的总平均回报率为 978%——与标普 500 指数相比，市场表现优于 211%。不要错过最新的前 10 名名单，该名单可与 Stock Advisor 一起使用，并加入由个人投资者为个人投资者构建的投资社区。
**Stock Advisor 的回报率截至 2026 年 5 月 30 日。 *
JPMorgan Chase 是 Motley Fool Money 的广告合作伙伴。Harsh Chauhan 对所提及的任何股票都没有持有任何头寸。The Motley Fool 在……中持有头寸，并推荐 Advanced Micro Devices、Intel、JPMorgan Chase 和 Nvidia。The Motley Fool 有一份披露政策。
本文中的观点和意见是作者的观点和意见，不一定代表 Nasdaq, Inc. 的观点和意见。
四大领先AI模型讨论这篇文章
"Nvidia's valuation already embeds aggressive growth assumptions, leaving little margin if AI spending growth decelerates or share is lost to custom silicon."
The article's $400 NVDA target rests on 87% EPS growth and expansion into server CPUs plus physical AI, yet it underplays how quickly custom ASICs from hyperscalers and AMD's MI300 series could erode Nvidia's 80-90% AI accelerator share. Forward P/E of 24x already prices in much of the data-center backlog; any Q3 guidance miss or macro-driven capex pause would compress multiples faster than new segments can offset. Physical AI revenue at a $9B run-rate remains tiny relative to the $200B long-term claim and faces years of regulatory and technical hurdles before scaling.
Nvidia's 95% sequential revenue guide and 75% gross margin already demonstrate pricing power that could let earnings exceed the $8.94 forecast, justifying a re-rating above 40x if AI capex stays elevated.
"The $400 thesis is mathematically plausible but depends entirely on Nvidia sustaining 85%+ revenue growth and 44x P/E multiples for 12 months—a combination that has rarely held when challenged by competition or macro slowdown."
The $400 target rests on a 2x P/E premium to the S&P 500 (44x forward multiple). That's aggressive but not impossible given 87% EPS growth guidance. However, the article conflates guidance with certainty. Nvidia's 80-90% AI accelerator dominance is real, but the $20B Vera CPU forecast and $200B physical AI TAM are speculative. The bigger issue: at current valuations, this thesis requires near-flawless execution AND no competitive erosion. The article ignores that AMD and Intel are shipping competitive products now, not in 2030. Gross margins at 75% are already elevated; sustaining that while scaling new products is harder than it sounds.
If Nvidia's guidance proves conservative and it hits $9.50+ EPS with 50x multiples justified by AI dominance, $400 is conservative. Conversely, if even one major cloud provider (AWS, Azure, Google) successfully deploys custom silicon at scale, Nvidia's accelerator TAM shrinks 20-30% and the valuation unwinds fast.
"Nvidia's ability to maintain 75% gross margins is the primary risk factor, as it invites aggressive vertical integration from its largest data center customers."
The article’s $400 price target relies on a simplistic 'multiple expansion' thesis, assuming NVDA should trade at 44x forward earnings. While the 85% revenue growth is staggering, investors must scrutinize the sustainability of 75% gross margins. History suggests that hardware dominance invites aggressive competition and customer-led silicon development (e.g., Google's TPUs or Amazon's Inferentia). If hyperscalers shift toward internal hardware to mitigate NVDA's pricing power, the 'monopoly' premium will compress rapidly. Furthermore, the article conflates fiscal 2027 projections with current market conditions, ignoring the cyclical nature of semiconductor capital expenditure. NVDA remains a high-beta play; its valuation is tethered to AI infrastructure spending, which is currently at a fever pitch.
If Nvidia successfully transitions from a pure hardware vendor to a full-stack software and services platform, it could justify a software-like multiple that makes current valuation metrics look cheap in hindsight.
"Nvidia can deliver outsized earnings growth, but its upside to $400 relies on a persistent AI capex cycle and a stable premium multiple—both of which are not guaranteed."
Nvidia's AI leadership supports a strong growth narrative: Q1 revenue up 85% to $81.6B, EPS up 140%, and forward guidance of $91B next quarter, with gross margins around 75%. The Vera server CPU and expanding physical AI TAM could sustain a long run, justifying a premium multiple. Yet the upside hinges on a durable AI capex cycle and Nvidia maintaining an 80–90% share in AI accelerators. The missing context includes potential deceleration in hyperscale demand, rising competition from AMD/Intel, and regulatory/export controls affecting China, plus potential margin pressure from mix and supply costs. A $400 target requires ongoing growth plus continued multiple expansion, which isn’t guaranteed.
The strongest counter is that even with solid growth, the market may not sustain a 44x forward P/E; if AI demand cools or rates rise, multiple compression could erase most of the upside.
"CUDA moat could postpone competitive impacts long enough for near-term beats but heighten later downside."
Claude and Gemini both highlight near-term competition from AMD, Intel, and custom ASICs, yet this underestimates how Nvidia's CUDA software lock-in could delay share loss by 2-3 years. That delay might let earnings beat estimates enough to support the 44x multiple temporarily, but it also means any eventual erosion hits when growth slows, amplifying downside risk beyond what the $400 target implies.
"CUDA lock-in buys time but doesn't prevent a sharp multiple re-rating once custom silicon adoption crosses a visibility threshold."
Grok's CUDA lock-in thesis deserves scrutiny. A 2-3 year delay assumes hyperscalers prioritize software compatibility over capex savings; that's optimistic. Google, Meta, and Amazon have already shipped custom silicon profitably. The real risk: Nvidia's earnings beat the $8.94 forecast near-term, justifying 44x, but the multiple compresses hard once custom silicon adoption accelerates—not gradually, but in a single quarter when a major player reports 30% lower GPU capex. That cliff risk isn't priced into $400.
"Foundry capacity constraints at TSMC act as a structural buffer that protects Nvidia's margins even if hyperscaler custom silicon adoption accelerates."
Claude is right about the 'cliff risk,' but you are all ignoring the supply-side constraint. Nvidia’s 75% gross margin isn't just pricing power; it’s a byproduct of TSMC’s CoWoS capacity bottleneck. If hyperscalers pivot to custom silicon, they still rely on the same limited foundry capacity. This creates an artificial floor for Nvidia’s dominance. The $400 target ignores that even with lower market share, Nvidia’s absolute revenue remains capped by physical manufacturing limits, not just demand.
"The real risk to the $400 bull-case is not demand, but potential capacity expansion by foundries and hyperscalers that could erode Nvidia's moat and trigger rapid multiple compression, making the 44x forward P/E brittle."
Gemini's supply-bottleneck argument hinges on CoWoS capacity, but that's a logistic constraint, not a permanent moat. If foundry capacity expands and hyperscalers accelerate internal silicon, Nvidia's revenue upside could be capped sooner than the bull thesis expects, and multiples could re-rate quickly on a single data point of capex slowdown or competitive losses. In other words, the 'floor' from supply isn't a floor for earnings or valuation.
The panel has mixed views on NVDA's $400 price target, with concerns about competition, potential capex slowdown, and the sustainability of high gross margins. While some panelists acknowledge NVDA's strong growth narrative and AI leadership, others warn of potential downside risks and the need for near-flawless execution.
Nvidia's CUDA software lock-in could delay share loss by 2-3 years, potentially allowing earnings to beat estimates and temporarily support the 44x multiple.
Rapid erosion of Nvidia's AI accelerator share due to competition from AMD, Intel, and custom ASICs, potentially leading to a cliff risk in earnings and multiple compression.