By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
Panelists agree that NVDA's valuation is tied to AI capex growth expectations, which may not persist, and Apple's services growth is at risk from antitrust scrutiny. They are neutral on both stocks, with varying degrees of confidence.
Risk: Potential margin compression for NVDA due to architectural disruption and AI demand slowdown, as well as antitrust risks for Apple's services growth.
Opportunity: None explicitly stated.
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 →
On July 17, Apple (NASDAQ: AAPL) overtook Nvidia (NASDAQ: NVDA) to reclaim the title of the largest company in the world. However, after Apple's strong run, Nvidia may once again be the better stock to buy.
Let's take a closer look at each stock to decide.
Where to invest $1,000 right now? Our analyst team just revealed what they believe are the 10 best stocks to buy right now, when you join Stock Advisor. See the stocks »
Apple has had a strong year, with its stock up more than 22%. The outperformance comes despite the company dealing with higher input costs from components like memory. However, in response, it has significantly raised prices for both hardware and some of its services that include cloud storage.
While the higher device prices could help lift revenue, there is a risk that it could also impact upgrade cycles. According to a poll by 9to5Mac, more than 90% of those surveyed said the higher prices would impact their buying habits, with nearly 40% saying they'd upgrade less often. That does add a potential risk at a time when the stock is sitting near all-time highs.
Apple's valuation has also risen. The stock now trades at a forward P/E of 34 times fiscal 2027 analyst estimates. It has frequently traded at a one-year forward P/E of around 24 times over the past several years, and this is its highest valuation during that stretch.
That said, Apple does have one of the best compounding business models out there. Once a user buys one of its devices, they tend to get locked into its ecosystem and use more of its high-margin services, including cloud storage, Apple Pay, and subscriptions.
One of its biggest revenue and profit contributors, meanwhile, is its revenue-sharing deal with Alphabet, which makes Google the default search engine on Apple's devices. This brings in more than $20 billion in revenue that falls straight to operating profit.
From purely a valuation standpoint, Nvidia now looks like the much more attractive stock than Apple. It trades at a forward P/E of only 16 times analyst estimates for fiscal year 2028 (ending January 2028), and it has continued to grow its revenue and profits at a breakneck pace.
The company is the dominant player in AI infrastructure with its graphics processing units (GPUs), especially when it comes to large language model (LLM) training. Nvidia has established a wide moat in this area due to its CUDA software platform, which it smartly gave away to universities and research labs that were doing early work on AI. As a result, most foundational AI code has been written on its software platform for its GPUs, which is why it is poised to remain the leader in AI model training.
However, the company has not sat still. Nvidia has a premier networking portfolio that now makes up the plumbing for its complete end-to-end server offerings. It's also developed its own ARM-based central processing units (CPUs), which are increasingly in demand due to the rise of agentic AI.
Nvidia also wisely "acquired" Groq this year and has integrated its language processing units (LPUs) into its CUDA ecosystem. LPUs use on-chip SRAM (static random-access memory) and are particularly beneficial during the decode phase of inference, letting Nvidia offer servers specifically for this task. With the inference market eventually expected to surpass the market for LLM training, Nvidia is well positioned.
The biggest question surrounding the stock is how long the AI build-out will last. Given the commentary from hyperscalers (owners of large data centers) and foundry Taiwan Semiconductor Manufacturing, there appears to be no let-up in AI infrastructure spending anytime soon, with demand remaining insatiable.
Nvidia is the cheaper stock with the better growth outlook. While there is a risk that the AI infrastructure build-out will eventually lose steam, it still looks like it is in the earlier innings. As such, it is the megacap tech stock I'd prefer to buy.
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 $364,562! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,247,668!
Now, it’s worth noting Stock Advisor’s total average return is 894% — a market-crushing outperformance compared to 207% 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 July 22, 2026. *
Geoffrey Seiler has positions in Alphabet. The Motley Fool has positions in and recommends Alphabet, Apple, Arm Holdings, Nvidia, and Taiwan Semiconductor Manufacturing. 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 valuation discount to Apple is illusory once growth normalizes; neither is clearly the better buy at current levels."
The article frames Nvidia as the superior megacap buy versus Apple on valuation (16x FY2028 vs AAPL's 34x FY2027) and AI-driven growth, while downplaying Apple's ecosystem lock-in and $20B+ Google search revenue. However, NVDA's forward multiple still embeds extreme expectations: analysts project ~40% CAGR through 2028. Apple's slower but durable 10-15% growth plus services margin expansion (now >70% of profit) looks less vulnerable to cyclical shocks. Missing context: hyperscaler capex commentary can shift rapidly; Taiwan Semi's upbeat tone reflects current backlog, not 2027 reality. NVDA trades at a premium to historical growth-adjusted peers once inference margins normalize.
The strongest case against favoring Apple is that Nvidia's CUDA moat and end-to-end AI stack (networking, Grace CPU, Groq LPUs) could sustain 50%+ revenue growth far longer than skeptics expect, making the 16x multiple appear cheap if inference spending explodes as projected.
"Nvidia's 'cheap' forward valuation is a trap built on unsustainable growth assumptions, while Apple's premium reflects a transition from hardware-dependent revenue to sticky, high-margin services."
The article's reliance on 2027/2028 forward P/E ratios is speculative at best, ignoring the cyclical nature of semiconductor demand. While Nvidia's 16x forward multiple looks 'cheap,' it assumes sustained hyper-growth that ignores potential margin compression as hyperscalers develop custom silicon (ASICs) to bypass Nvidia's tax. Apple, conversely, is being mispriced as a hardware company rather than a high-margin services ecosystem. Trading at 34x forward earnings is steep, but the recurring revenue from services provides a defensive moat that Nvidia's capital-expenditure-dependent model lacks. I am neutral on both; the market is currently over-optimizing for AI infrastructure while under-appreciating the volatility of Apple's consumer upgrade cycle.
The bull case for Nvidia rests on the 'AI infrastructure as a utility' thesis, where the build-out is a multi-decade capital cycle rather than a temporary spike, rendering current valuation multiples irrelevant.
"NVDA's valuation looks cheap only if you believe the AI build-out sustains current capex intensity indefinitely; AAPL's premium is high but defensible given recurring revenue and ecosystem lock-in, making this a risk-tolerance call, not a valuation arbitrage."
The article's valuation comparison is misleading. Yes, NVDA trades at 16x forward P/E versus AAPL's 34x, but that gap reflects fundamentally different risk profiles, not a bargain. NVDA's 16x assumes sustained AI capex growth; if hyperscaler spending normalizes to historical levels (say, 20-25% of revenue vs. current 30%+), NVDA re-rates sharply lower. AAPL's 34x premium reflects $20B+ annual Google payments (nearly 100% margin), installed base lock-in, and proven pricing power. The 9to5Mac poll is noise—Apple has survived every 'upgrade cycle death' prediction since 2014. Real risk: NVDA faces architectural disruption (custom silicon from Meta, Google, Amazon) and margin compression in inference, not valuation alone.
If AI capex remains insatiable for 5+ years as the article claims, NVDA's moat (CUDA, end-to-end stacks) justifies 16x even with 30% annual growth, while AAPL's services dependency on Google creates regulatory/antitrust tail risk the article barely mentions.
"Nvidia's premium multiple hinges on an unsustainably long AI demand cycle; a slowdown could trigger meaningful valuation compression."
While Nvidia looks cheap on forward earnings, the thesis rests on an endless AI capex cycle that may not persist. Nvidia’s data-center demand has ridden years of hyperscale spend, but macro cooling, memory price normalization, and potential saturation of inference workloads could compress growth. Competitive risk from AMD/Intel accelerators, as well as shifts in developer ecosystems away from CUDA or toward open frameworks, looms. The article’s confident stance ignores the possibility of margin compression and a near-term multiple re-rate if AI demand slows. Apple, by contrast, offers a durable cash-flow machine with services and ecosystem optionality that could re-rate independently of AI cycles.
NVDA’s CUDA moat and continued data-center spending could prove more durable than it appears; if AI capex remains robust and margins stay elevated, the stock can re-rate despite near-term headwinds.
"Antitrust hit to AAPL's Google revenue poses larger immediate derating risk than NVDA inference margin pressure."
Claude correctly flags NVDA's architectural disruption risk but underweights how quickly inference margins are already compressing in disclosed gross margin trends. Nobody has tied the $20B Google payments directly to AAPL's antitrust exposure: a potential breakup or revenue share cut could instantly derate that 34x multiple by 20-25%. That's the sharper regulatory tail risk than cyclical capex normalization.
"Apple's antitrust risk is structural to its ecosystem, not just a revenue-share issue, making its premium multiple fundamentally unstable."
Grok, your focus on Google revenue is valid, but you are all missing the real antitrust catalyst: Apple's 'walled garden' is under direct DOJ scrutiny regarding hardware-software integration, which is far more existential than a Google payment haircut. If Apple is forced to open iOS, that 34x multiple on services growth evaporates. Meanwhile, Nvidia's risk isn't just custom silicon; it’s the potential for a 'utility' pricing model shift as GPU supply finally meets demand, crushing margins.
"Apple's regulatory risk is payment renegotiation, not iOS breakup; Nvidia's margin compression is a near-term earnings headwind, not a valuation mirage."
Gemini's iOS breakup scenario is theoretically catastrophic but politically implausible—Apple's walled garden is a feature, not a bug, and DOJ has struggled to articulate a coherent breakup remedy. More pressing: Grok's Google payment haircut (20-25% multiple compression) is *imminent* if DOJ forces revenue-share renegotiation, which has concrete precedent. That's a 2-3 year tail risk, not speculative. Nvidia's margin compression from inference is already visible in gross margin guidance; that's a 12-month headwind, not a 2028 problem.
"Google payments tail risk is not a certain 2–3 year derating; hyperscaler architectural shifts could erode NVDA's CUDA moat and trigger a faster margin/growth re-rate than assumed."
Claude, your near-term margin headwind is sensible, but you overplay the Google payments risk as a sure 2–3 year derating. DOJ timing is uncertain, and even if renegotiated the impact could be offset by price leverage or software monetization. The bigger, under‑discussed risk is a potential architecture shift by hyperscalers that erodes CUDA’s moat—compression of both growth and margins could re-rate NVDA faster than the article assumes.
Panelists agree that NVDA's valuation is tied to AI capex growth expectations, which may not persist, and Apple's services growth is at risk from antitrust scrutiny. They are neutral on both stocks, with varying degrees of confidence.
None explicitly stated.
Potential margin compression for NVDA due to architectural disruption and AI demand slowdown, as well as antitrust risks for Apple's services growth.