Temporary pullbacks can provide investors with a chance to buy shares at a discount relative to recent highs. When a company’s long-term growth story remains intact, a short-term decline may have little to do with its underlying business performance. 

Instead, the selloff may simply reflect shifting market sentiment or broader market conditions. For patient investors, these periods can offer attractive entry points into companies they may have previously considered too expensive.

NVIDIA Remains a Core AI Growth Story

After dropping from about $235 to about $200, the tech giant has become technically oversold on relative strength, MACD, and Williams’ %R. We also have to remember that NVIDIA (NASDAQ: NVDA) is the backbone to the AI boom, which shows no signs of cooling off.

For one, the global artificial intelligence market is valued at approximately $617 billion. Projections indicate the market size will exceed $3.4 trillion by 2033, fueled by widespread adoption and rapid scaling of generative and agentic AI, as noted by Grand View Research. 

Two, the rapid growth of data centers further supports the long-term outlook. There are currently around 4,000 operational data centers in the United States, with another 1,500 to 3,000 either planned or under construction. Research also shows strong regional expansion across the country, particularly in the South and Midwest. Globally, there are now more than 10,000 data centers. Those numbers are only set to climb.

Oracle Is Showing Signs of Recovery

After getting beaten up on earnings, Oracle (NYSE: ORCL) is starting to pivot higher, too.

EPS of $2.11 beat by 15 cents. Revenue of $19.2 billion, up 20.8% year over year, beat by $110 million. While solid, the stock cell after it forecast higher capital expenditure for fiscal 2027 than analysts’ estimates and said it expects to raise about $40 billion in debt and equity in fiscal 2027 to support its capital investment program.

However, according to Citi analysts, the stock is a buy. “It was a mixed print for ORCL as Q4 saw a solid beat, with bookings upside, accelerating IaaS growth, and decreased financing intensity, but underwhelming FY27 EPS guidance on gross margin pressure. We still see more positives than negatives,” said analysts led by Tyler Radke.

“While gross margins are under pressure from significant capacity ramps in FY27, we believe the Q4 bookings strength, reduced financing needs (from pre-payments/BYOC) give increasing confidence to the ramp in long-term targets. We further raise our Infrastructure revenue forecasts, with our FY27 EPS staying largely unchanged but FY29- 30 moving slightly higher,” said Radke and his team, as quoted by Seeking Alpha.

Lowe’s Could Benefit From Improving Housing Trends

After a devastating pullback from about $350 to about $290 in May, shares of Lowe’s (NYSE: LOW) are starting to come back strong, too.

Last trading at $332.81, we’d like to see it rally back to $350 initially.

For one, Citi, for example, just upgraded Lowe’s to a buy rating following the latest pullback, arguing that the worst may already be priced into the stock.

According to Citi analysts, the housing and home improvement market could see gradual improvement throughout 2026. While growth may remain modest, there is still significant pent-up demand for home improvement spending after years of elevated mortgage rates and reduced housing turnover. 

“We see 2026 as a year of gradual improvement, even if the growth is slightly lower. There is pent-up demand for home improvement spending on a multi-year basis as existing home sales step up to higher levels and lower rates drive increased engagement with projects,” said the firm, as quoted by Seeking Alpha.

Two, it typically benefits from hurricane season, which runs from June 1 to late November. This segment is “naturally positively exposed to preparation and recovery efforts,” says Morgan Stanley.  These typically see a boost in sales post-storm as damaged property is repaired.”

Undervalued Stocks Can Lead to Opportunities

While volatility can feel uncomfortable, it often creates some of the best opportunities for investors who are willing to look beyond the noise. For investors, the key is not trying to perfectly time every move, but rather identifying quality companies when sentiment temporarily pulls prices down.

CrowdStrike: Investors Are Selling The Future, Not The Business

CrowdStrike (NASDAQ: CRWD) dropped sharply after reporting revenue growth of 26%, EPS growth of 51%, record operating cash flow of $384 million, record free cash flow of $279 million, and raised full-year guidance — a combination that reads like exactly what investors claim to want, which makes the reaction worth understanding rather than simply accepting at face value.

The frustration has less to do with endpoint security and more to do with what comes next. Management keeps talking about a second act built around autonomous systems and AI-driven security operations, and investors who had already priced in early confirmation of that transition sold when the quarter delivered strong current-period results without material evidence that autonomous security is generating meaningful revenue yet. 

After rallying from roughly $350 in March to nearly $780 before earnings, the stock had become one of the market’s biggest winners. Expectations were stretched. When the company delivered another strong quarter rather than a transformative one, investors took profits.

Yet the longer term picture remains constructive. Shares remain well above the 50-day moving average near $521 and the 200-day near $479, which is not where a stock trades when the money that matters has decided to leave. The market is questioning how valuable CrowdStrike’s next growth engine becomes. It is not questioning the quality of what’s already built — and that distinction is where the opportunity lives.

Cadence Design Systems: The Toll Booth Nobody Talks About

Most investors debating the AI race spend their time arguing over which chip company wins. Cadence Design Systems (NASDAQ: CDNS) gets paid before that race even begins.

Every advanced semiconductor requires sophisticated design software before it can be manufactured, tested, or deployed at scale. And as AI models grow larger and more computationally demanding, the chips required to run them grow more complex in direct proportion, making Cadence’s electronic design automation tools more critical to players like Nvidia (NASDAQ: NVDA), Advanced Micro Devices (NASDAQ: AMD), Broadcom Inc (NASDAQ: AVGO), Marvell Technology (NASDAQ: MRVL) and others in the ecosystem.

The stock surged from roughly $270 in April to more than $410 before pulling back toward $380. A correction after a move like that should surprise nobody. What matters is where the stock corrected to. Shares remain comfortably above the 50-day moving average near $336 and the 200-day near $325, and volume during the decline never suggested the kind of institutional distribution that precedes a genuine breakdown. The chart looks like a stock digesting a powerful breakout, not beginning a downtrend.

AI companies compete with each other. Chipmakers fight over market share. Cloud providers battle for dominance. Every single one of them still needs increasingly sophisticated chip designs to pursue any of those ambitions. The market sees a software company. I see a toll booth collecting from every lane of the AI highway simultaneously — and toll booths don’t lose relevance when traffic increases.

Arista Networks: The Infrastructure Play Everyone Keeps Overlooking

Investors love talking about GPUs. They spend far less time discussing what happens after the GPUs arrive. What happens is that data has to move, constantly, at enormous scale, between servers that need to communicate faster than the chips themselves can process. That is where Arista Networks (NYSE: ANET) sits, providing the networking infrastructure that allows massive AI clusters and data centers to function at the performance levels the workloads actually demand. As AI models grow larger and clusters scale wider, networking becomes more important rather than less, because moving data efficiently between compute nodes increasingly determines overall system performance. 

The recent selloff dragged Arista lower alongside names facing entirely different structural risks, as investors conflated AI infrastructure spending concerns with businesses that sit in fundamentally different positions within that spending chain. Shares pulled back from approximately $180 toward the $150 area, but remain above both the 50-day moving average near $153 and the 200-day near $141, with the longer-term uptrend structurally intact and the recent weakness resembling consolidation rather than conviction selling.

The market is treating Arista as another AI stock exposed to the same risks as the names driving the narrative. The business is considerably more durable than that framing suggests, because as long as data centers keep expanding, networking is not discretionary spending that gets deferred when budgets tighten. It is the infrastructure that makes everything else function.

The Setup Across All Three

CrowdStrike’s business remains strong while the market debates its second act. Cadence keeps collecting from every company building the next generation of chips regardless of who wins the competition between them. Arista sits inside one of the most critical bottlenecks in modern computing infrastructure with a position that strengthens as the workloads it supports grow more demanding.

The selloff created uncertainty. But for investors willing to look past the short-term narrative and focus on what the businesses are actually doing, it may also have created the entry that the run-up never offered.

“We presume that fundamentals and related price action tell us we’re in middle innings. In turn, both trailing and forward PEs should be expected to moderate going forward, thus increasing the burden on earnings growth and related expectation to drive index price action.”

The AI Boom Shows No Signs of Slowing Down

With the global AI market already surpassing $230 billion in 2024, analysts now see a clear path to multi-trillion-dollar expansion—and the next five years may deliver the strongest gains yet.

Forecasts now place AI’s value between $1.7 and $3.5 trillion by the early 2030s, with the most aggressive estimates topping $7 trillion by 2035. And judging by the surge in corporate investment, the market is moving toward the high end of those projections.

Plus, we have to consider that artificial intelligence will continue to create massive demand for data centers, which will lead to further demand for NAND. After all, NAND is a vital part of the AI infrastructure for massive amounts of data storage, speed and performance. We also have to consider that demand for more data centers (which will need more NAND) are exploding.

The data tells the story

According to MIT Technology Review, there are about 3,000 data centers across the U.S. Plus, according to a report from McKinsey, $5.2 trillion in AI infrastructure investments will be needed by 2030. Again, growing demand for data centers will mean growing demand for more NAND memory in an already tight market.

In addition, “As the demand for data processing and storage continues to surge, data centers are experiencing unprecedented growth. McKinsey’s analysis ‘suggests that demand for AI-ready data center capacity will rise at an average rate of 33 percent a year between 2023 and 2030 (reflecting a trend that is already underway.),’” according to analysts at McKinsey, as reported by BOMA International.

The “AI Bubble” Argument Is Fading

Warnings of an “AI bubble” are increasingly being dismissed by top analysts.

Goldman Sachs says, “it believes the AI story is just getting started – and the investments that seem huge today will be dwarfed by the benefits AI will deliver,” as noted by Quartz.com. 

Long term, the investment bank says that AI adoption could add $20 trillion to the U.S. economy. AI, according to Goldman Sachs, is already delivering those gains in productivity when deployed right.”

JPMorgan’s Mary Callahan Erdoes added, “AI is presenting opportunities not fully appreciated or understood yet,” as noted by CNBC. “AI itself is not a bubble. That’s a crazy concept… We are on the precipice of a major, major revolution in a way that companies operate.”

How Investors Can Tap Into the AI Boom

One way is to invest in heavyweights like Nvidia (NASDAQ: NVDA) and Advanced Micro Devices (NASDAQ: AMD).

Or, if you prefer to diversify, consider ETFs such as:

One of the Leading AI-Focused ETFs

If you want to diversify at a lower cost, there are ETFs like the Global X Artificial Intelligence & Technology ETF (NASDAQ: AIQ). With an expense ratio of 0.68%, the ETF invests in companies that may benefit from the continued development and use of artificial intelligence (AI). Some of its top holdings include Palantir, Oracle, Broadcom, Netflix, Nvidia, Microsoft, and Meta Platforms, to name a few of its 86 total holdings. 

A Core ETF for Automation and AI Expansion

Another top AI ETF to consider is the Global X Robotics and Artificial Intelligence ETF (NASDAQ: BOTZ). With an expense ratio of 0.68%, the ETF invests in companies that should benefit from the increased adoption of robotics and AI. Some of its 49 holdings include Nvidia, Keyence, DynaTrace, SMC Corp., Intuitive Surgical, Upstart Holdings, and C3.ai.

One reason is the growing excitement around Marvell. Jensen Huang, CEO of Nvidia (NASDAQ: NVDA), just said Marvell could be the next potential trillion-dollar company. He highlighted the importance of Marvell’s networking and connectivity chips, which help data centers move information quickly between thousands of connected processors. His comments helped drive a rally in Marvell and other optical networking stocks.

Another reason is that optical networking has become a critical technology for artificial intelligence. Unlike traditional copper cables, optical networks transmit data using pulses of light, allowing for much higher speeds, greater bandwidth, and lower delays. As AI systems grow larger and more complex, the need for faster data transfer continues to increase.

AI Networking Boom Continues

According to Goldman Sachs, networking is becoming the next major area of investment in AI infrastructure. Faster, low-latency data exchange is essential for improving AI computing performance and supporting increasingly powerful models.

The firm estimates that the optical networking market could grow from roughly $15 billion in 2026 to nearly $154 billion by 2028.

In addition, we have to consider that the industry is facing a significant supply-demand imbalance. As artificial intelligence companies build larger computing clusters, they need tens of thousands of GPUs connected through ultra-fast optical links. At that scale, traditional copper cables can no longer deliver the required performance.

To help, Nvidia has reportedly asked suppliers to increase production capacity for indium phosphide lasers by as much as 20 times by 2030. The company has also committed approximately $2 billion to help secure future supply.

As artificial intelligence adoption accelerates, companies that provide the optical networking technology needed to connect massive systems could be positioned for substantial growth.

Early Days for Artificial Intelligence

For one, consider this. There are about 4,000 operational data centers in the U.S. right now. An additional 1,500 to 3,000 are being planned or under construction. According to Pew Research, the South has 754 planned data centers. The Midwest has 419 planned. The West has 277 planned, and the Northeast has about 106 planned. 

Two, the artificial intelligence boom is still accelerating.

Forecasts now place the value of artificial intelligence between $1.7 and $3.5 trillion by the early 2030s, with the most aggressive estimates topping $7 trillion by 2035. And judging by the surge in corporate investment, the market is moving toward the high end of those projections.

Three, warnings of an “AI bubble” are increasingly being dismissed by top analysts.

Goldman Sachs says, “it believes the AI story is just getting started – and the investments that seem huge today will be dwarfed by the benefits AI will deliver,” as noted by Quartz.com. 

Long term, the investment bank says that artificial intelligence adoption could add $20 trillion to the U.S. economy. AI, according to Goldman Sachs, is already delivering those gains in productivity when deployed right.”

JPMorgan’s Mary Callahan Erdoes added, “AI is presenting opportunities not fully appreciated or understood yet,” as noted by CNBC. “AI itself is not a bubble. That’s a crazy concept… We are on the precipice of a major, major revolution in a way that companies operate.”

“So, if you say to yourself, is AI in a bubble, I feel you have to get very granular on how you’re going to answer that, because in the U.S., we’re starting to gain traction, but we’re nowhere near the ability to have the stuff all to the bottom line.” 

For investors, that means the upside potential is still there.

To provide a quick summary of the bullish case for NVDA stock, it comes down (obviously) to artificial intelligence. No matter how much criticism and controversy AI generates, the reality is that AI has fundamentally shifted our paradigm. If you’re not on board with the AI narrative, you’re hopelessly lost. And since NVIDIA sells differentiated compute, it’s arguably less likely to stumble under commoditization concerns.

However, this line of reasoning represents thesis risk, which is very much limited for NVIDIA — and some might argue nonexistent. What investors and traders should be concerned about is equity risk. This is a different category and requires a nuanced, analytical approach.

Too often, when it comes to popular securities like NVDA stock, financial writers wax poetic about AI spending, infrastructure demand and projected runways. Well, I’m sorry, but it makes me wonder why publishers even pay money for such garbage contributions. These are narratives that are clearly baked into the story.

What serious analyst is declaring that AI is a fad and demand will collapse next quarter? No one — and that’s the risk. When you have an environment where people are constantly bidding up the bullish narrative and giving little thought to the bearish case, unnecessary exuberance can easily embed itself in the target security.

Fundamentally, the risk isn’t that no one believes in the AI narrative undergirding NVDA stock; no, it’s that everybody believes it. As such, there is a premium associated with the enthusiasm. Essentially, the spread between expectation and eventual reality must be robust enough to justify this premium.

It’s quite possible, then, to have a situation where the thesis pans out but the equity doesn’t cooperate. It’s not that the thesis was wrong (because it wasn’t). Rather, the thesis didn’t quite cover the differential between expectations and reality.

Why You Need to Pay Attention to the Smart Money’s Take on NVDA Stock

Interestingly, the smart money has a very different approach to your typical buy-and-hold retail perspective — and that’s evidenced by the volatility skew. By definition, the skew identifies implied volatility (IV) across the strike price spectrum of a given options chain. Since IV reflects the pricing potential of a security at the selected strike, traders have an incentive to cover the implied move.

Basically, the skew can be considered an insurance market. On any given day, a popular security will go up or it will go down. Professional traders, especially those responsible for massive funds, must probabilistically determine which trajectory is more likely. Subsequently, the hedging transactions distort the skew, allowing observers to better determine smart money sentiment.

What’s fascinating about the further-out July 17 expiration date is that the volatility skew for Nvidia stock shows relative put dominance across the strike price spectrum. The most obvious point is that below the spot price, put IV rises sharply (and swings above calls at the lowest depths). This framework represents classic crash insurance behavior; that is, the market is assigning a meaningful premium to downside protection.

Does that mean that a crash is imminent for NVDA stock? Not at all — the smart money shouldn’t automatically be considered prescient. However, we give these traders the label “smart” because of their sophisticated transactional acumen.

I would also argue here that these pros are intellectually honest about NVDA stock. While they recognize that Nvidia can march higher (judging by the rising call IV), put dominance is still the order of the day for strikes above spot.

Conclusion? The smart money apparently believes NVIDIA stock is fairly priced relative to current expectations. And I think that’s the right take. When you look at the same skew for something like Micron Technologies (NASDAQ: MU), traders are pricing for upside convexity all the way, with seemingly little prioritization for downside protection.

That makes me nervous. NVIDIA stock also makes me pensive at this hour, but at least the smart money is recognizing the risk.

Going Conservative on NVIDIA Stock

Naturally, the disagreement about NVDA stock isn’t about whether it’s a good opportunity in the abstract. It’s up roughly 59% in the trailing year; it’s a good opportunity. But the question is about the magnitude of goodness. Again, it’s the differential between expectations and reality.

I’m going to ride with the smart money — and the data would suggest the same. If you swing back using a dataset to 1990, a random 10-week-long position would be expected to generate a forward median distribution landing between $208 and $235 (assuming a starting price of $214.86).

Under the current quantitative structure, in the past 10 weeks, NVIDIA stock printed six up weeks, leading to an upward slope across the period. Under this specific 6-4-U sequence, you would expect the forward 10-week distribution to land between $208 and $240.

So, the math is quite simple. Comparing aggregate to conditional sequences, you would statistically (from an observational standpoint) be looking at no added risk. But on the reward side, the distribution shifts positively, with the right-side tail differential hitting 2.13%.

Now, is 2.13% that much of a difference compared to a randomly aggregated 10-week-long position? Since the reward penalty is 0%, there is a modestly positive asymmetry here. However, it’s not a pound-the-table type of play.

This is where a very conservative bull call spread might be in order. Probably the most ambitious spread I would be looking at is the 220/225 bull spread expiring July 17. With this trade, I’m looking for NVDA stock to rise through the $225 strike at expiration. If it does, the maximum payout clocks in at nearly 144%.

From machine learning and automation to generative AI tools and advanced semiconductors, the artificial intelligence market is expected to expand rapidly over the next several years, creating major opportunities for both companies and investors.

According to Grand View Research, the global AI market could expand from about $137 billion in 2022 to more than $1.8 trillion by 2030. Other studies suggest AI may add trillions of dollars to the global economy over the next decade as companies use it to improve productivity, lower costs, and increase profits.

Because of this massive potential, many investors want exposure to AI-related stocks. Investors are increasingly viewing artificial intelligence as a long-term megatrend similar to the early growth of the internet or cloud computing. That excitement has helped fuel significant gains in many AI stocks, but it has also increased volatility as valuations rise and competition intensifies across the sector. For newer investors, especially, finding a balanced way to participate in the future of AI without taking excessive risk has become an important part of building a long-term portfolio.

Companies like NVIDIA (NASDAQ: NVDA) and Advanced Micro Devices (NASDAQ: AMD) have become popular choices because they help power AI systems through advanced computer chips and data center technology. However, choosing the right AI stocks can be difficult. The industry changes quickly, competition is intense, and even promising companies can experience large swings in price.

For investors who want a simpler and potentially safer way to invest in the future of AI, exchange-traded funds (ETFs) can be a smart option. AI ETFs allow investors to own a collection of companies connected to artificial intelligence instead of relying on a single stock. This diversification can reduce risk while still providing exposure to the fast-growing AI market.

Why Investors Are Looking Beyond Individual AI Stocks

Global X Artificial Intelligence & Technology ETF

One ETF many investors consider is the AIQ, also known as the Global X Artificial Intelligence & Technology ETF (NASDAQ: AIQ). This fund focuses on companies that could benefit from the development and use of AI technologies across different industries. The ETF has an expense ratio of 0.68%, making it a relatively affordable way to gain broad exposure to the AI sector.

The fund includes major technology companies and AI leaders such as Palantir Technologies (NASDAQ: PLTR), Oracle (NYSE: ORCL), Broadcom (NASDAQ: AVGO), Netflix (NASDAQ: NFLX), Microsoft (NASDAQ: MSFT), and Meta Platforms (NASDAQ: META). By investing in multiple companies, the ETF spreads risk across different parts of the AI industry, including cloud computing, software, semiconductors, and digital platforms.

Global X Robotics and Artificial Intelligence ETF 

Another one of the top AI ETFs to consider for the future of AI is the Global X Robotics and Artificial Intelligence ETF (NASDAQ: BOTZ). With an expense ratio of 0.68%, the ETF invests in companies that should benefit from the increased adoption of robotics and AI. Some of its 49 holdings include NVIDIA (NASDAQ: NVDA), Keyence (OTCMKTS: KYCCF), DynaTrace (NYSE: DT), SMC Corp. (OTCMKTS: SMCAY), Intuitive Surgical (NASDAQ: ISRG), Upstart Holdings (NASDAQ: UPST), and C3.ai (NYSE: AI).

ETFs Can Help You Capture the Future of AI

Artificial intelligence is expected to remain one of the fastest-growing sectors in the global economy for years to come. As companies continue investing in machine learning, automation, robotics, and AI-powered software, investors have an opportunity to benefit from the industry’s long-term expansion. While individual AI stocks like NVIDIA and Advanced Micro Devices may continue to attract attention, AI-focused ETFs can provide a diversified approach to investing in this rapidly exploding market.

For long-term investors, AI ETFs may represent one of the most practical ways to participate in the future of AI while managing risk. Instead of depending on a single company to dominate the industry, ETFs allow investors to benefit from broader trends across semiconductors, cloud infrastructure, software, robotics, and automation. As artificial intelligence adoption expands throughout the global economy, diversified AI ETFs could continue to attract investors seeking both growth potential and a more balanced approach to one of the market’s most transformative technologies.

The headline numbers in the company’s Q1 2027 earnings report showed adjusted EPS of $1.87, as revenue exploded 85% year over year to $81.61 billion. Analysts had expected the company to earn an adjusted $1.75 per share on $79.19B in revenue.

Data center revenue came in at $75.2 billion, as compared to estimates of $73.48 billion.  Edge Computing revenue was $6.4 billion, up 29% year-over-year. Adjusted gross margin came in at 75% for the quarter, in line with estimates. NVDA also generated $48.55 billion.

Moving forward, NVIDIA expects revenue to be $91 billion, plus or minus 2%. The company also said that it does not expect any revenue from China for the period. Analysts had expected the company to generate $87.3 billion in revenue.

Even more impressive, NVIDIA also added $80 billion to its share buyback program. And it raised its quarterly dividend to $0.25 per share, up from $0.01 per share, payable to shareholders of record on June 4 and June 26.

And Yet, That Just Wasn’t Good Enough for Wall Street

Analysts say investors “have got used to NVIDIA delivering stellar results and amid some concerns that it will face growing competition,” as noted by the BBC.  “NVIDIA represents 8% of the S&P 500. Unless there’s a belief in this continued parabolic growth, it’s difficult for investors to get super excited, although NVIDIA posted outstanding numbers.”

Another key concern is China.

During the earnings call, the company said that, “While the U.S. government has approved licenses for H200 to be shipped to China-based customers, we have yet to generate any revenue, and we are uncertain whether any imports will be allowed into the country. As a result, consistent with last quarter, we are not including any China data center compute revenue in our outlook.”

In addition, despite crushing estimates and delivering strong growth, NVIDIA’s high expectations are becoming harder to justify as investors grow more cautious. Not helping, investors are concerned that NVIDIA may not be able to maintain its extremely fast revenue growth forever. After several quarters of huge gains driven by AI demand, some investors fear growth could start slowing as comparisons become tougher and the market matures.

Plus, as noted by CEO Jensen Huang, the company has “largely conceded” China’s AI chip market to Huawei. “The demand in China is quite large,” Huang told CNBC. “Huawei is very, very strong. They had a record year, they’ll likely, very likely, have an extraordinary year coming up, and their local ecosystem of chip companies are doing quite well, because we’ve evacuated that market. “We’ve really largely conceded that market to them.”

The Good News – AI Isn’t Slowing

Warnings of an “AI bubble” are increasingly being dismissed by top analysts.

Goldman Sachs says, “it believes the AI story is just getting started – and the investments that seem huge today will be dwarfed by the benefits AI will deliver,” as noted by Quartz.com. 

Long term, the investment bank says that AI adoption could add $20 trillion to the U.S. economy. AI, according to Goldman Sachs, is already delivering those gains in productivity when deployed right.”

JPMorgan’s Mary Callahan Erdoes added, “AI is presenting opportunities not fully appreciated or understood yet,” as noted by CNBC. “AI itself is not a bubble. That’s a crazy concept… We are on the precipice of a major, major revolution in a way that companies operate.”

“So, if you say to yourself, is AI in a bubble, I feel you have to get very granular on how you’re going to answer that, because in the U.S., we’re starting to gain traction, but we’re nowhere near the ability to have the stuff all to the bottom line.” 

For investors, that means the upside potential is still there.

Yet, NVIDIA Corporation (NYSE: NVDA) reported Q1 earnings for FY27 with a revenue of $81.6 billion and diluted EPS of $0.76 instead. That contradiction was the most fascinating part of the quarter because Nvidia did not merely overcome a headwind that could have crippled most companies; it produced numbers so large they almost buried China’s story.

And the deeper I went, the harder it became to view NVIDIA as “just” a semiconductor company going forward.

China Has Crippled Growth Stories Before

Apple Inc. (NASDAQ: AAPL) regularly faces scrutiny whenever Chinese sales slow. Nike Inc. (NYSE: NKE) spent years treating China as a critical growth engine before slowing demand became a recurring concern. Starbucks Corporation (NASDAQ: SBUX) spent decades building China into its most important international growth engine. Yet, increasing local competition has pressured the company to explore strategic alternatives for its China business as growth slowed

That’s how global markets normally work. When a major market weakens, growth slows, and investors reassess expectations.

NVIDIA faced something far worse than slowing demand.

The company disclosed that H20 export restrictions resulted in a $4.5 billion charge during the quarter. It also disclosed an additional $2.5 billion in H20 revenue it could not ship due to those restrictions.

That is $7 billion of impact connected to a single product line. For most companies, a disruption of that magnitude would dominate the quarter. For NVIDIA, it became background noise.

NVIDIA Showed How Little China Matters

The quality of numbers NVIDIA released this quarter is disturbing, but in a good way.

Revenue reached $81.6 billion. Gross profit reached $58.8 billion. Operating income reached $44.1 billion. Operating cash flow reached $48.8 billion. Free cash flow reached $26 billion.

The board also approved an additional $80 billion share repurchase authorization.

When you compare this to the $7 billion connected to China again, NVIDIA still generated enough cash in one quarter to fund entire industries.

But make no mistake, this story isn’t about how NVIDIA survived China’s restrictions. In fact, focusing on that alone could cause a fatal misinterpretation of these earnings, which could lead you to a false conclusion about the AI boom and where it’s headed.

What I’m trying to tell you is that the global AI spending has expanded so rapidly it absorbed the restrictions that broke down the internet a couple of months ago.

And nowhere was that reality more visible than inside the Data Center business. Let me explain.

What Selling The Capacity Behind An Infrastructure Race Looks Like

A step further in this report, you’d bump into Data Center revenue climbing to $75.2 billion, up 92% year-over-year.

Not only that, Gaming also generated $3.8 billion. Professional Visualization generated $509 million. Automotive generated $567 million.

Place those figures next to each other, and the transformation becomes impossible to ignore.

The old NVIDIA still exists, which is where most investors’ theses are stuck.

But the new NVIDIA completely dominates it. The company no longer looks like a chipmaker benefiting from AI demand. It looks like the company supplying the computational backbone behind one of the largest infrastructure buildouts in modern history.

The quarter repeatedly pointed toward the same destination; Blackwell systems ramped. AI factories expanded. Sovereign AI projects accelerated. Inference demand continued climbing.

The world’s largest technology companies, the government and enterprises… Everyone is still spending because, before an AI model can reason, before a robot can navigate a warehouse, before an autonomous vehicle can interpret its surroundings, someone must build the computing infrastructure first. And NVIDIA is sitting at the center of it all.

A Powerful Uptrend

NVDA exploded higher following earnings, confirming what had already become one of the strongest charts in the market. The stock recently broke above the key $200 resistance zone, turning a level that capped rallies for months into potential support.

Technically, NVDA remains firmly above its 20-, 50-, and 200-day moving averages, signaling a strong momentum across multiple timeframes.

The earnings-driven surge also pushed the stock toward the upper boundary of its rising channel near $235-$240. While some short-term consolidation would be normal after such a sharp run, the trend remains firmly in the bulls’ favor as long as NVDA holds above the $200 breakout area.

At the moment, buyers continue treating every pullback as an opportunity to gain exposure to the AI infrastructure buildout.

The Spending Still Hasn’t Hit The Wall

Of course, Wall Street keeps searching for signs that AI spending will cool.

NVIDIA’s guidance told a different story.

The company guided for approximately $91 billion in Q2 revenue despite the continued impact from export restrictions.

That may be the most important figure in the entire report.

Because guidance arrives after management has already seen customer orders, deployment schedules, capacity plans, and infrastructure demand.

And yet the company still expects another leap higher. This is why NVIDIA is losing billions of dollars tied to China, while still producing revenue, profits, cash flow, and guidance that most corporations could not generate under ideal conditions…could go down as the most critical AI story in decades.

And if this infrastructure race keeps accelerating faster than global tension, the company’s biggest challenge may no longer be finding customers. It may be keeping up with them.

The artificial intelligence boom is still accelerating, with no signs of slowing down. 

 With the global AI market already surpassing $230 billion in 2024, analysts now see a clear path to multi-trillion-dollar expansion, creating substantial opportunity for investors.

Within the global market, one of the fastest-growing segments is the conversational AI market, for example. Projected to hit roughly $41 billion to $62 billion in market value by 2032, it could surpass $136 billion by 2035, according to Research and Markets. 

The rapid growth is being fueled by rising corporate adoption of AI for customer engagement, automation, personalized experiences, and operational efficiency improvements. Advances in voice recognition, generative AI, and omnichannel communication platforms are also accelerating adoption across multiple industries.

Why the Conversational AI Market Is Growing So Quickly

As noted by Research and Markets, “The surge in AI-powered customer support is a primary growth driver, as businesses shift from traditional call centers to efficient, personalized systems handling high-volume inquiries around the clock.”

The research firm went on to note that major companies are investing in solutions that not only respond but also execute actions, accelerating market expansion through enhanced efficiency and user satisfaction. Here are three stocks to consider if you want exposure to this expanding market.

SoundHound AI Is Expanding Its Voice Commerce Ecosystem

SoundHound AI (NASDAQ: SOUN) recently announced a new partnership with OpenTable, a global leader in restaurant technology, to launch a fully conversational in-vehicle voice AI reservations agent – as part of SoundHound’s in-car voice commerce platform. 

The new AI-powered system allows drivers and passengers to search for restaurants, check availability, and book reservations entirely through voice commands integrated into their vehicle infotainment systems.

The platform connects users to a network of more than 60,000 restaurants worldwide, streamlining what would otherwise be a multi-step process into a seamless conversational AI experience.

In addition, partnerships spanning Red Lobster, Burger King UK, Peet’s Coffee, Duke Health, Lucid Motors, and BNP Paribas are helping position SoundHound as one of the more diversified pure-play conversational AI companies in the market.

RingCentral Is Integrating AI Into Business Communications

RingCentral (NYSE: RNG) is the global leader in AI-powered business communications. The company recently highlighted growing adoption of its RingEX platform.

“Over a million users rely on RingCentral RingEX as a lightweight contact center, enabling employees to respond to customers alongside everyday work,” said Kira Makagon, President and COO of RingCentral.

“The future is customer-centric — uniting AI, unified communications, and contact center capabilities.”

As more companies adopt AI-enhanced communication tools to improve customer engagement and efficiency, RingCentral may benefit from expanding enterprise demand for conversational AI solutions.

Roundhill Generative AI ETF Offers Diversified AI Exposure

With an expense ratio of 0.75%, the Roundhill Generative AI & Technology ETF (NYSEARCA: CHAT) is the world’s first generative AI ETF. Some of its 38 holdings include NVIDIA (NASDAQ: NVDA), Alphabet (NASDAQ: GOOGL), Meta Platforms (NASDAQ; META), Microsoft (NASDAQ: MSFT), Oracle (NYSE: ORCL), Palantir Technologies (NASDAQ: PLTR), and Alibaba Group Holding (NYSE: BABA). 

These companies help power AI infrastructure, cloud computing, semiconductors, enterprise software, and generative AI applications, supporting the continued expansion of conversational AI technologies.

For investors seeking broader exposure to the AI sector instead of individual conversational AI stocks, CHAT may offer a diversified approach.

Conversational AI Stocks Could Benefit From Long-Term Enterprise Adoption

As artificial intelligence continues reshaping industries worldwide, conversational AI is emerging as one of the market’s strongest long-term growth opportunities.

Businesses are rapidly adopting AI-driven customer engagement tools, voice-enabled systems, and automation technologies to improve efficiency and customer experiences. Companies positioned at the center of this transition could benefit from years of sustained demand growth.

Whether through innovative platforms like SoundHound AI, enterprise communication providers like RingCentral, or diversified exposure through the Roundhill Generative AI & Technology ETF, investors have multiple ways to participate in what could become one of the largest technology expansion trends of the decade.

That was a good quarter. Then management disclosed a single figure that completely changed how I read the quarterly report.

Specifically, customers prepaid Tower Semiconductor $290 million tied largely to silicon photonics demand, while the company locked in approximately $1.3 billion in contracted silicon photonics revenue for 2027.

But nobody hands a semiconductor foundry $290 million years ahead because things are “comfortable”. That only happens when industries start realizing future access may become more important than future pricing.

AI Boom Is Beginning To Slam Into Physics

For the last two years, the AI race revolved around one thing: more GPUs. But President Trump’s latest visit to China, alongside Nvidia CEO Jensen Huang and other semiconductor executives, revealed something bigger than “GPUs” underneath the AI race.

That is, semiconductor infrastructure has now matured into full geopolitical infrastructure. Governments, hyperscalers, and chip companies are no longer fighting only over AI models or computing power. The pressure is shifting deeper into the systems moving electricity, networking, bandwidth, and data fast enough to keep future AI infrastructure scaling.

This is why I wasn’t surprised when NVIDIA (NASDAQ: NVDA) launched silicon photonics networking systems designed to move data using light rather than traditional electrical signaling, as AI infrastructure keeps scaling up and becoming more power-intensive.

Or when Broadcom (NASDAQ: AVGO) reported that it is also expanding co-packaged optics infrastructure built specifically for AI networking scalability. Because the real problem now is no longer just compute. The systems themselves are beginning to choke on: heat, electricity, latency, bandwidth, and the physical challenge of moving enormous amounts of data between GPUs fast enough to keep scaling AI workloads efficiently. 

And if you think that means optical communication systems are set to become mandatory for next-generation AI infrastructure as copper-based systems approach physical scaling limits, you’d be right.

Actions Speak Louder Than Voice

Tower Semiconductor’s operating profit surged roughly 96% year-over-year to approximately $65 million while operating margin expanded from 9.1% to 15.7%. Management also guided Q2 revenue toward roughly $455 million, which would mark the highest quarterly revenue in company history.

Those figures were strong. But the prepayments were stronger because customers usually do not reserve future foundry capacity this aggressively unless they believe the next bottleneck is already forming underneath the market.

In fact, as we speak, Tower Semiconductor is set to quintuple monthly silicon photonics output by the end of 2026 after securing large AI-linked customer agreements tied directly to future infrastructure demand.

That combination stayed with me the entire earnings call: $290 million upfront, $1.3 billion in future contracts, record guidance, exploding margins, and aggressive optical infrastructure expansion underneath the AI buildout.

The quarter started feeling less like a semiconductor earnings report and more like companies quietly scrambling to secure positions before the next infrastructure bottleneck gets worse. Let that settle in for a second.

How Big Is The $290 Million Prepayment?

TSEM surged more than 22% after earnings, closing near $270 after exploding higher from the prior close of around $220. Volume also surged above 6 million shares, several times higher than the stock’s normal trading activity.

Wall Street does not usually gap a semiconductor foundry higher by that magnitude overnight unless investors believe something inside the business has changed.

Buyers aggressively repriced the stock after management disclosed: $290 million in customer prepayments, $1.3 billion in contracted silicon photonics revenue, and plans to expand photonics capacity to meet accelerating AI infrastructure demand.

Technically, the earnings move also shattered the upper trend channel that had capped price action for months while pushing TSEM dramatically above the 20-day moving average near $216 and the 50-day moving average near $185. The gap itself completely altered the chart’s structure.

The key zone now sits around roughly $240-$245 near the top of the breakout area. If institutional buyers continue defending that earnings gap aggressively, investors may begin treating TSEM less like a specialty foundry and more like a strategic infrastructure company sitting underneath the next phase of AI scaling.

I’m In Position with Tower Semiconductor

If you still think the AI race revolves around whichever company sells the fastest chips, you’ll need to catch up. The fight is already moving lower into the infrastructure stack itself.

AI systems now need faster networking, lower latency, lower heat generation, and dramatically more efficient data transfer just to keep scaling larger workloads. That pressure is pushing the industry toward optical communication systems capable of replacing parts of the copper-based architecture that modern AI systems increasingly struggle to scale efficiently.

And Tower Semiconductor suddenly sits right in the middle of that transition.

Customers already prepaid $290 million before much of that future infrastructure even fully arrives. That doesn’t look like normal semiconductor demand to me.

Instead, it looks like an industry starting to realize the next AI bottleneck may already be forming underneath the surface – and quietly paying up before everybody else catches on.