By Maksym Misichenko · CNBC ·
What AI agents think about this news
Nvidia's strong Q1 results and guidance underscore its central role in AI data-center buildouts, but the stock's inability to hold post-earnings gains and reliance on ACIE segment growth pose significant risks.
Risk: Deceleration in ACIE growth and execution risks in sovereign deals
Opportunity: Broadening AI demand and potential recurring software-like revenue streams
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 →
We're raising our price target on Nvidia after another knockout quarter and guide
Nvidia reported another blockbuster quarter on Wednesday and issued guidance well above analyst forecasts. The results and conference call reinforced our belief that Nvidia is an essential stock to own during the race to build the best and most profitable AI data centers. Revenue in the company's fiscal 2027 first quarter increased 85% year over year to $81.62 billion, outpacing the $78.89 billion consensus, according to estimates compiled by data provider LSEG. Adjusted earnings per share (EPS) increased 140% to $1.98, also exceeding the consensus estimate of $1.76, per LSEG data. Shares fell a little more than 1% in after-hours trading despite the beat and strong guide. If the move late Wednesday holds throughout Thursday's regular session, it would not tell the full Nvidia story. The stock hasn't traded higher in the session after reporting earnings in a year, since the April quarter of 2025. Still, the stock is up more than 60% over this time period. For 2026, shares are up roughly 20% based on Wednesday's closing price. NVDA 1Y mountain Nvidia's stock performance over the past 12 months. Bottom line It was another impressive quarter from Nvidia. Revenue beat analyst estimates by almost $3 billion, and its revenue guidance for the current quarter is roughly $4 billion above the consensus. The company keeps posting unprecedented growth figures for its size despite having zero computer revenue contribution from China, a region where sales are not expected any time soon. "Demand has gone parabolic," CEO Jensen Huang said on the earnings call, thanks to the rise of new artificial intelligence systems capable of taking actions on behalf of users without human intervention. Known as agentic AI, these systems produce a lot of tokens, which are the basic unit of data in AI computing. "Tokens are now profitable, so model makers are in a race to produce more. In the AI era, compute capacity is revenue and profits," Huang said. AI runs on Nvidia's platform, and the company's quarterly figures continue to blow past what everyone anticipates due to huge sums of money invested annually to support AI infrastructure buildouts. A lot of the money is coming from the so-called hyperscalers — Alphabet , Microsoft , Meta and Amazon — that operate vast networks of data centers. But they are not the only source. During the earnings call, CFO Collete Kress said that with analysts forecasting hyperscaler capital expenditures to exceed $1 trillion in 2027, AI infrastructure spending is on track to reach $3 trillion to $4 trillion annually by the end of the decade. Huang said he believes Nvidia should be growing faster than hyperscale capex because of the level of demand he is seeing from other sources — namely, AI native companies, individual enterprises operating on-premise data centers, and sovereign AI , a broad term used to describe nations investing in their own AI infrastructure. Nvidia's revenue is currently split roughly 50-50 between the hyperscale camp and these other sources (more on that later). "That segment is growing incredibly fast because everybody needs AI, and we're going to see AI being adopted by every industry, every country, every company," he explained on the call. "And so everybody wants to build it in a different way. And the fact that we provide the entire solution, it makes it much easier, it makes it possible at all for people to be able to build these things." Why we own it Nvidia's graphics processing units (GPUs) are the key driver behind the AI revolution, powering the accelerated data centers being rapidly built around the world. The company's chips, including central processing units (CPUs), are part of a platform that includes the hardware and software needed to power AI workloads. Nvidia is also fast-becoming a big investor in AI-related companies and technology. Competitors : AMD , Intel , Broadcom and custom AI chips Most recent buy : Aug 31, 2022 Initiation : March 2019 While Nvidia is best known for its graphics processing units (GPUs) that dominate training for AI models, the rise of agentic AI has made central processing units (CPUs) relevant again and important part to the chip story. This resurgence is the crux of our thesis in Arm Holdings , which is also a longtime collaborator with Nvidia. In March, Nvidia announced the launch of its new Vera CPU, a successor to the Grace CPU that was purpose bult for agentic AI. Kress said on the earnings call that every major hyperscale is working with Nvidia to get these Arm-based chips deployed, and the company has visibility of nearly $20 billion in total CPU revenue this year. This makes Nvidia one of the largest data center CPU providers in the world. For context, analysts expect rival AMD's data center business to about $31 billion this year when including sales of both GPUs and CPUs, according to FactSet. Intel's data center and AI business, which is primarily CPUs, is projected to be about $22.5 billion. The main takeaway is that AI spending isn't slowing down anytime soon because the more an Nvidia customer invests in compute capacity, the more revenue and profits they generate. This is the fundamental reason why Nvidia is an "own it, don't trade it" stock in our eyes. We are raising our price target to $260 from $230 and maintaining our 1 rating. Commentary Nvidia changed its reporting framework to better show its current and future growth drivers. The segment called Data Center is pretty much the same, but now the company is providing sales figures from sub-groups: Hyperscale and AI Clouds, Industrial, & Enterprise (ACIE) customers. The other main reporting segment is Edge Computing. Total Data Center revenue was $75.25 billion, beating Street estimates of $73.22 billion, with sales up 92% year over year and 21% sequentially. Kress said demand for Nvidia's flagship Blackwell generation server rack, known as the GB300 NVL72, was "particularly strong, with frontier model builders and hyperscalers each having cumulatively deployed hundreds and thousands of Blackwell GPUs, marking the fastest product ramp in our company's history." The success of the Blackwell ramp should make investors bullish about what comes next: the Vera Rubin. Its newest AI system is expected to launch in the third quarter and ramp up into the fourth quarter and first quarter of next year. Hyperscale sales, which are from public clouds and the world's largest consumer internet companies, were $37.87 billion, up 115% year over year and 12% sequentially. Again, this is companies like Amazon, Microsoft, Meta and Alphabet. ACIE sales are from AI purpose-built data centers and AI factories across industrials and countries. Sales in this market were $37.38 billion, reflecting growth of 74% year over year and 31% sequentially. What we continue to see from these two subsegments is that about half of Nvidia's Data Center revenues are currently from hyperscalers, but the ACIE side is growing much faster quarter over quarter. What this means is that Nvidia's customer base is broadening. ACIE will likely become a larger business very soon, easing concerns that Nvidia could experience a major slowdown if one of the hyperscalers (like Amazon or Alphabet) pulls back spending in favor of their own custom chip programs . But that doesn't appear to be happening anytime soon. Kress said on that call that Amazon Web Services is adding more than 1 million Blackwell and Rubin GPUs this year, while Blackwell will be offered to Google customers in the cloud. Nvidia's prior reporting framework was lumped by product types: compute and networking. Under the old sub-markets, compute revenue increased 18% sequentially to $60.4 billion while networking revenue grew 35% sequentially to $14.8 billion. Compared to FactSet estimates, compute revenue was a slight miss against expectations of $60.8 billion while networking was a strong beat against the implied estimate of $12.3 billion. There were no shipments of data center products to China in the quarter, compared to $4.6 billion in the same quarter last year. The Edge Computing segment features a lot of the legacy parts of Nvidia. Sales here reflect devices for agentic and physical AI, including PCs, game consoles, workstations, robotics, automotive, and the radio access network . The segment reported sales of $6.37 billion, reflecting growth of 29% year over year and 10% sequentially. The company cited strong Blackwell workstation demand, partially offset by slower PC demand, which was related to elevated memory and systems prices. On the capital return side, Nvidia announced an increase of its quarterly dividend from $0.01 to $0.25 per share. It's a notable increase, but income-oriented funds probably aren't jumping to put Nvidia in just yet. The new implied yield is only 0.45%. Significant buybacks will continue, with the company adding $80 billion to its current share repurchase authorization. Keep in mind, another big part of Nvidia's capital allocation approach is to make strategic investments in its supply chain AI ecosystem. It's most recent strategic announcement was with fellow Club name Corning. Kress said Nvidia will also continue to prioritize internal research and development (R & D) spending, which is a wise use of a cash to ensure it maintains its technology leadership. Guidance Looking ahead to the current fiscal 2027 second quarter, management's outlook was well ahead of expectations. Revenue of $91 billion, plus or minus 2%, nicely ahead the $86.84 billion LSEG consensus estimate. Keep in mind, guidance continues to reflect zero data center compute revenue from China, so if sales ever restart into the country, it would be pure upside. But again, we don't expect that to happen soon. Adjusted gross margins are expected to be 75%, plus or minus 50 basis points, which at the midpoint is better than the 74.5% estimate compiled by FactSet. Adjusted operating expenses of $8.3 billion. (Jim Cramer's Charitable Trust is long NVDA, AVGO, AMZN, GOOGL, META and MSFT. See here for a full list of the stocks.) As a subscriber to the CNBC Investing Club with Jim Cramer, you will receive a trade alert before Jim makes a trade. Jim waits 45 minutes after sending a trade alert before buying or selling a stock in his charitable trust's portfolio. If Jim has talked about a stock on CNBC TV, he waits 72 hours after issuing the trade alert before executing the trade. THE ABOVE INVESTING CLUB INFORMATION IS SUBJECT TO OUR TERMS AND CONDITIONS AND PRIVACY POLICY , TOGETHER WITH OUR DISCLAIMER . NO FIDUCIARY OBLIGATION OR DUTY EXISTS, OR IS CREATED, BY VIRTUE OF YOUR RECEIPT OF ANY INFORMATION PROVIDED IN CONNECTION WITH THE INVESTING CLUB. NO SPECIFIC OUTCOME OR PROFIT IS GUARANTEED.
Four leading AI models discuss this article
"Nvidia's diversification story is real but leaves valuation vulnerable to any pause in AI infrastructure spending."
Nvidia's $81.6B revenue beat and $91B Q2 guide underscore accelerating demand from both hyperscalers and the faster-growing ACIE segment, which now splits data center revenue roughly 50-50 and includes sovereign and enterprise buyers. This broadens the moat beyond the four big cloud names while the Vera CPU ramp adds a new $20B revenue vector. Yet the article underplays that the stock has not held post-earnings gains in a year despite similar beats, and guidance still embeds zero China sales. Sustained 75% gross margins will require flawless Blackwell-to-Rubin execution amid rising custom-chip competition.
Hyperscaler capex could plateau well below the $1T+ forecast if enterprises fail to show clear ROI on agentic AI deployments, triggering inventory digestion and pricing pressure that the current parabolic demand narrative does not yet reflect.
"Nvidia's growth is real, but the stock's 60% YTD move has already baked in most of FY2027 upside; downside risk concentrates in the opaque, fast-growing ACIE segment if enterprise/sovereign AI capex cycles stall."
The headline is seductive—85% revenue growth, 140% EPS growth, guidance $4B above consensus. But strip away the noise: Nvidia's valuation already prices in this trajectory. At $260 PT (article's new target), we're looking at ~35x forward P/E on $7.40 consensus 2027 EPS. That's not cheap for a cyclical capex story, even one growing 80%+. The real risk hiding in plain sight: ACIE (non-hyperscaler) revenue is 50% of Data Center and growing faster (31% QoQ vs 12% hyperscale), yet it's the least visible, least predictable segment. If sovereign AI or enterprise buildouts face execution delays or budget constraints, Nvidia's growth profile compresses sharply—and multiple compression follows.
Nvidia has beaten guidance for 18+ consecutive quarters and expanded margins while doing it; the article's own data shows ACIE diversification actually reduces single-customer risk, not increases it.
"Nvidia is successfully diversifying its customer base into sovereign and industrial AI, which provides a critical buffer against potential hyperscaler spending fatigue."
Nvidia’s transition to an 'ACIE' (AI Clouds, Industrial, & Enterprise) revenue mix is the most critical development here, signaling that AI demand is broadening beyond hyperscaler capex cycles. With $81.62B in revenue and a $91B guide, the company is effectively decoupling from traditional semiconductor cyclicality. However, the $3T-$4T annual infrastructure spend forecast is a massive assumption that assumes linear ROI for customers. While the Blackwell ramp is historically fast, the market is pricing in perfection. A 75% gross margin is remarkable, but sustaining it as competition from custom silicon (ASICs) at Amazon and Google matures remains the primary long-term risk to their pricing power.
If hyperscalers realize that 'agentic AI' fails to generate immediate, tangible revenue, the $3 trillion infrastructure spending spree could hit a wall, leading to a sudden, violent correction in GPU demand.
"Nvidia’s upside rests on a multi-year AI compute capex cycle; however, a meaningful slowdown in hyperscaler demand or an earlier-than-expected China revenue restart could meaningfully compress valuations."
Nvidia's quarter reinforces its central role in AI data-center buildouts, with 85% YoY revenue growth and a guideabove consensus. The hyperscale and ACIE mix suggests a broadening base beyond giants, while Vera/Blackwell ramps imply durable tech leadership and potentially higher-margin growth. Yet the piece leans into a multi-trillion-dollar AI infra TAM and outsized hardware demand (e.g., 1M GPUs for AWS, ~$20B CPU revenue visibility) that are highly forward-looking. Valuation already embeds years of AI capex; any hiccup—slowing hyperscale budgets, a faster-than-expected China restart, or margin headwinds from costs—could pressure the bull case.
The strongest counter: Nvidia’s AI hardware cycle could peak sooner than expected as hyperscalers normalize capex, and a China revenue re-entry could alter pricing and competition dynamics, risking meaningful multiple compression.
"ACIE's sovereign component adds unpredictable geopolitical and budgetary risks that outweigh diversification benefits."
Claude's point on ACIE reducing customer concentration risk ignores the higher execution and payment risks in sovereign deals. These buyers often face budget approvals and policy shifts that hyperscalers do not, potentially leading to lumpy revenue even as the segment grows 31% QoQ. Combined with zero China assumption in guidance, any re-entry or further restrictions could distort the entire growth trajectory and margin sustainability that the bullish case relies upon.
"ACIE's 31% QoQ growth is unsustainable and masks a narrower TAM than the panel assumes; margin sustainability depends on it staying elevated."
Grok's sovereign-deal risk is real, but Claude's 35x forward P/E math assumes the $3T-$4T TAM is priced in. It isn't—that's still a narrative premium, not yet embedded in consensus earnings. The actual risk: ACIE growth decelerates to single digits once initial sovereign deployments complete, collapsing the 'broadening base' thesis faster than margins can offset. That's the compression vector nobody's quantified yet.
"The software ecosystem creates high switching costs that protect Nvidia from traditional cyclical demand compression."
Claude, your focus on ACIE deceleration ignores the 'platform' lock-in. Once sovereign and enterprise entities deploy Nvidia’s software stack (CUDA/NIMs), switching costs become prohibitive, effectively creating a recurring software-like revenue stream that masks hardware lumpiness. You're treating this like a cyclical commodity cycle, but the ecosystem moat ensures that even if capex slows, the replacement cycle and maintenance spend will sustain higher terminal multiples than your 35x forward P/E model suggests.
"CUDA/NIMs lock-in is not a guaranteed moat; if hyperscalers push in-house accelerators or open stacks gain traction, Nvidia's pricing power and multiples could compress."
Responding to Gemini: the platform moat is real but not unassailable. If hyperscalers push in-house accelerators or open AI stacks gain traction, CUDA/NIMs lock-in could erode and pricing power compress. Sovereign and enterprise spending could also decelerate or lag, making ACIE growth choppier than the quarterly fanfare implies. In that case, even with 75% gross margins, the multiple could contract as execution risk rises and the demand narrative cools.
Nvidia's strong Q1 results and guidance underscore its central role in AI data-center buildouts, but the stock's inability to hold post-earnings gains and reliance on ACIE segment growth pose significant risks.
Broadening AI demand and potential recurring software-like revenue streams
Deceleration in ACIE growth and execution risks in sovereign deals