By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel consensus is bearish, warning of a potential downturn in AI infrastructure stocks due to cyclical nature of semiconductor industry, memory ASP collapse, and overinflated valuations that embed aggressive growth.
Risk: Margin compression due to ASP drop and potential slowdown in hyperscaler capex
Opportunity: Durable earnings power from power infrastructure plays like Bloom Energy, if grid constraints are addressed
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 →
Beyond the hyperscaler AI developers are the infrastructure companies that supply the components that make it work.
Since these are mostly smaller companies with niche product lines, they're experiencing explosive growth due to high demand.
As the world embraces artificial intelligence (AI), and companies that provide development infrastructure are thriving, AI stocks are driving higher market gains. As a group, the industry is outperforming the market. One standard signal is the Nasdaq-100, which features trending tech stocks and is beating the S&P 500 this year. Many exchange-traded funds (ETFs) that focus on AI are doing even better. Consider the Roundhill Generative AI and Technology ETF, the iShares U.S. Power and Infrastructure ETF, and the Pacer Data and Infrastructure Real Estate ETF.
Will AI create the world's first trillionaire? Our team just released a report on the one little-known company, called an "Indispensable Monopoly" providing the critical technology Nvidia and Intel both need. Continue »
It's only May, but several hot artificial intelligence stocks have already more than doubled. Take a look at Bloom Energy (NYSE: BE), Sandisk (NASDAQ: SNDK), Lumentum (NASDAQ: LITE), Micron Technology (NASDAQ: MU), and Intel (NASDAQ: INTC).
Here's why they can still go higher.
Bloom Energy makes stand-alone energy servers using its proprietary fuel cell technology that can power organizations off the electric grid. It has been in operation for 25 years, serving a wide range of industries, from retail to space. However, it's exploding in the AI era, as it can be easily deployed and scaled to power high-performance data centers.
It has been reporting phenomenal performance. In the first quarter, revenue inreased 130% year over year, and management raised full-year guidance from a 60% increase to an 80% increase. Operating cash flow turned positive, increasing $184.3 million to $73.6 million. Bloom has been unprofitable for most of its life, and this sales jump is finally pushing it into profitable territory, with $70 million in net income in the quarter.
Sandisk has been the AI story of the year as its NAND flash memory products become more in demand for data storage. While AI companies need a wide range of memory products, SanDisk's specific technology can retain data without a continuous power source.
Revenue has been accelerating as demand soars and prices rise. In the 2026 fiscal third quarter (ended April 3), revenue increased 251% year over year, driven by a 233% increase in data center revenue. Operating income increased 319% to more than $4 billion.
So far, management is expecting this to continue. It's guiding for revenue to grow sevenfold in the fourth quarter, and for adjusted earnings per share (EPS) of around $30, up from a $0.30 loss per share last year.
Lumentum makes optical products that are crucial for high-speed data centers. It's one of the infrastructure providers that help move large amounts of data over broadband networks, connecting servers at a lower cost.
Like the other companies on the list, Lumentum had a completely different business servicing other technologies, like medical equipment and manufacturing, but data centers are driving the business right now.
In the 2026 fiscal third quarter (ended March 28), revenue increased 90% year over year, and the company reported a generally accepted accounting principles (GAAP) operating profit, up from a loss last year. CEO Michael Hurlston said, "As our key growth drivers of co-packaged optics and optical circuit switches begin to kick in, we would expect further increases in earnings power."
Micron also makes memory chips, and it's been in this business for a long time. Many types of technology rely on memory chips to operate, such as computers and smartphones, and Micron makes DRAM memory chips that power short-term memory and high-bandwidth memory (HBM) chips. It also sells NAND memory chips.
Since these are necessary components of AI chips, the business has been skyrocketing. Revenue increased 70% year over year in the 2026 fiscal second quarter (ended Feb. 26), and earnings per share rose from $4.60 to $12.07.
The company is benefiting from high demand and low supply, and management expects that to continue in the short term.
Intel had a massive technology business prior to the AI boom, but it began to fade into irrelevance when it failed to anticipate and respond to the changing tech landscape. Tech giants moved from Intel's central processing units (CPUs) to graphics processing units (GPUs) and other AI chips to drive AI development, and revenue was falling while margins contracted.
However, CPUs are in high demand again as they're necessary for high-inference workloads. Intel has made some important deals with large AI companies, and it's getting back to growth. Intel is backed by decades of tech leadership, and this could be a strong turnaround story in the coming months and years.
Before you buy stock in Bloom Energy, 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 Bloom Energy 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 $469,293! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,381,332!
Now, it’s worth noting Stock Advisor’s total average return is 993% — 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 May 17, 2026. *
Jennifer Saibil has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Bloom Energy, Intel, Lumentum, and Micron Technology. 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
"Memory and legacy CPU suppliers face sharp mean-reversion risk once supply normalizes, contrary to the article's linear growth narrative."
The article spotlights AI infrastructure names like Bloom Energy, Lumentum, Micron, and Intel after they doubled on data-center demand, citing 70-130% revenue spikes and raised guidance. Yet it skips semiconductor cycle history where memory ASPs collapse once supply catches up, Intel's repeated node delays, and Bloom's long unprofitability despite the recent $70M quarter. Outdated tickers such as SNDK and 2026 fiscal references also undermine credibility. Momentum ETFs could amplify near-term moves, but valuations already embed aggressive growth that leaves little margin if hyperscaler capex moderates.
Continued multi-year AI buildout by the largest cloud providers could keep component shortages intact, allowing these stocks to re-rate higher even from current levels if earnings keep beating.
"These stocks have priced in multi-year hypergrowth; the real risk isn't whether AI demand exists, but whether current guidance assumes an unrealistic absence of any demand normalization."
The article conflates two distinct narratives: hyperscaler AI demand (real, durable) and infrastructure beneficiaries (real but cyclical). SanDisk's 251% YoY revenue growth and 7x Q4 guidance are extraordinary, but the article doesn't address whether this reflects genuine demand acceleration or inventory restocking ahead of a potential slowdown. Bloom Energy's $70M net income swing is material, but turning profitable after 25 years on one quarter of tailwinds is fragile. The article also omits critical context: memory chip cycles are notoriously boom-bust, and these valuations already price in sustained hypergrowth. Micron's 70% revenue growth is impressive, but supply-constrained markets eventually normalize.
If AI capex moderates even 15-20% from current levels—a plausible scenario given macro uncertainty and rising rates—these companies face not just slower growth but potential demand destruction, since they're selling discretionary infrastructure, not essential components. Valuation multiples suggest zero margin for disappointment.
"The article contains fundamental factual errors, including citing a non-public company (SanDisk) and projecting 2026 fiscal data, rendering its investment thesis unreliable."
This article is fundamentally flawed and potentially deceptive; it cites 'SanDisk' as a publicly traded stock, despite Western Digital acquiring it in 2016. Furthermore, it references 2026 fiscal data, which suggests either a hallucinated timeline or a major error in reporting. While the AI infrastructure thesis—specifically regarding power constraints for data centers—is valid, the execution here is reckless. Bloom Energy (BE) is indeed a beneficiary of the 'power-hungry' AI narrative, but investors should be wary of the massive capital expenditure required to scale fuel cell deployments. Relying on this list for investment signals is dangerous given the blatant factual inaccuracies regarding company status and fiscal reporting.
If we ignore the glaring factual errors, the core thesis holds that infrastructure providers are the 'picks and shovels' of the AI gold rush, offering a safer entry point than the overextended hyperscalers.
"The touted AI upside in these five stocks is likely to fade as cyclicality, margin pressure, and execution risk outweigh speculative AI demand."
The piece is bullish on a handful of AI‑adjacent names and frames a multi‑bag rally as proof of an AI demand boom. But the argument rests on cherry‑picked headlines and cyclical growth, not durable earnings power. Bloom Energy pivots to AI‑era data centers, yet profits remain fragile and cash flow hinges on a revenue surge that may not sustain. SanDisk/Micron/Intel exposure sits in a memory/optics cycle that is notoriously volatile, with pricing pressure and capex swings ahead. The piece glosses over execution risk, valuation risk, and the concentration of AI upside among a few hyperscalers. If AI demand cools or supply chains flex, gains may fade quickly.
The strongest counterpoint is that the AI hype-driven rally in these five names could be a cyclic peak; if demand normalizes or margins compress, there’s little durable earnings power to support elevated multiples.
"Supply expansions by memory firms risk triggering ASP collapses within 18 months, amplifying downside across the AI infra group."
Gemini rightly flags the SanDisk and 2026 errors that erode trust in the piece. Yet the deeper flaw is assuming hyperscaler capex stays elevated without addressing how memory suppliers like Micron are already expanding capacity aggressively. Historical cycles show ASPs dropping 30-50% within 18 months of a supply ramp. This could compress margins faster than the AI narrative suggests, hitting all these names simultaneously rather than in isolation.
"ASP cycles are real, but the timing of supply normalization—not its inevitability—determines whether these stocks crater or re-rate higher first."
Grok's ASP compression timeline is the crux. But 18 months assumes symmetric supply ramp—historically false. Micron's capex cycles lag demand by 12-18 months; if hyperscalers front-load orders now, supply tightness persists through 2025. The real risk isn't margin compression per se—it's *when* it hits. If it's Q4 2025, these valuations crater. If it's Q2 2026, earnings still beat expectations through next earnings season, and momentum extends. Article's silence on capex timing is damning.
"The AI infrastructure bottleneck is shifting from semiconductor supply to power grid capacity, creating a more durable tailwind for energy-adjacent firms than for memory manufacturers."
Claude, you’re betting on a delayed supply response, but you’re ignoring the 'power wall' mentioned by Gemini. Bloom Energy and similar power-infrastructure plays aren't just cyclical; they are limited by physical grid capacity and permitting, not just chip yields. Even if Micron’s supply ramps, the data centers can't scale if the power isn't there. We are looking at a bottleneck shift from silicon to electrons, which makes the infrastructure thesis more durable than the memory cycle.
"Power constraints matter, but timing and policy shifts will determine whether the AI hardware rally persists; model multiple power-timing scenarios rather than assuming a one-way lift."
Responding to Gemini. The 'power wall' is a timing/policy risk, not an absolute limit. Grid constraints could slow builds, but shift capex toward on-site generation—a durable tailwind for Bloom Energy. The bigger risk is hyperscaler spend timing: if AI capex slows in late 2025, the rally could unwind even if power remains supportive, and investors should model multiple scenarios for power vs. timing outcomes.
The panel consensus is bearish, warning of a potential downturn in AI infrastructure stocks due to cyclical nature of semiconductor industry, memory ASP collapse, and overinflated valuations that embed aggressive growth.
Durable earnings power from power infrastructure plays like Bloom Energy, if grid constraints are addressed
Margin compression due to ASP drop and potential slowdown in hyperscaler capex