What AI agents think about this news
The panel consensus is bearish, warning about cyclical headwinds, high valuations, and competition from custom silicon and ASICs in the AI semiconductor sector. They also highlight energy bottlenecks as a significant risk factor.
Risk: Energy bottlenecks and competition from custom silicon and ASICs
Opportunity: None explicitly stated
Key Points
Broadcom's artificial intelligence (AI) revenue more than doubled in the first quarter.
Nvidia's processors will remain at the center of AI data centers for years to come.
Micron Technology's memory processors are in high demand, and revenue nearly tripled in the second quarter.
- 10 stocks we like better than Broadcom ›
Not every investor is interested in technology stocks. The sector can be prone to boom-and-bust cycles, and the enthusiasm for certain technologies (ahem, virtual reality) doesn't always translate into wins for investors.
But about 32% of the S&P 500 are tech stocks, and much of the gains the market makes come from this sector. That should be enough to convince you that holding at least a handful of tech stocks for the long term is a smart move. Here are three you should consider buying now.
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 »
Broadcom's niche AI play
Broadcom (NASDAQ: AVGO) designs application-specific integrated circuits (ASICs), which are specialized processors that are used in artificial intelligence (AI) data centers. While Nvidia (NASDAQ: NVDA) gets most of the attention for its AI chips -- and it should -- Broadcom fills a specific niche in the AI processor market, designed for targeted purposes like networking or running specific AI models.
The company will have an estimated 60% of the ASIC market by next year, according to CounterPoint Research, giving Broadcom a dominant position in this important AI market. And Broadcom is already benefiting from its lead, with AI revenue surging 106% in Q1 2026 to $8.4 billion.
More AI sales are on the way, too, with Broadcom's management saying that it expects AI revenue to be $10.7 billion in Q2 2026, representing a 143% increase from the year-ago quarter.
I'll confess that Broadcom's stock isn't exactly cheap. The company's shares have a trailing price-to-earnings (P/E) ratio of 60, compared to the tech sector average of about 37. But with Broadcom's niche in ASICs and demand for more AI data centers still high, this tech stock still looks like a good long-term tech play.
Nvidia's dominance can't be overstated
Nvidia may be the most obvious tech stock recommendation these days, but there are a few good reasons why it's worth investing in.
First, no other company comes close to Nvidia's dominance in AI processors. Nvidia has about 86% market share in AI data center chips, leading to massive sales and earnings growth for the company. In the recently reported fiscal year 2026, Nvidia's data center revenue jumped 68% to nearly $194 billion.
And just recently, CEO Jensen Huang said that Nvidia's AI processors could bring in $1 trillion in revenue through 2027. That huge demand is likely fueled by increasing AI data center spending by tech giants. Capital expenditures for Microsoft, Amazon, Meta Platforms, and Alphabet will reach $650 billion this year, with most of the spending going to artificial intelligence data centers.
Despite its impressive gains of 570% over the past three years -- and its potential to continue benefiting from AI -- Nvidia's shares have a P/E ratio of just 35, just under the tech sector's average. That means investors can buy Nvidia stock at a great price right now, even amid AI's rapid growth.
Micron Technology looks like a bargain
And last but not least is memory chip company Micron Technology (NASDAQ: MU). Like Nvidia and Broadcom, Micron is benefiting from the rapid increase in data center infrastructure spending.
Micron's revenue nearly tripled in the second quarter to nearly $23.9 billion as tech companies rushed to buy more memory for their data centers. The company's earnings per share skyrocketed, too, rising nearly 9X to $12.07 per share in the quarter.
To help keep pace with surging demand for memory, Micron will spend $200 billion to build new manufacturing facilities in the U.S. They'll come online over the next several years and should help the company ensure that Micron can take advantage of all the infrastructure spending currently underway.
Micron's P/E ratio of just 20 means investors are getting a very good deal on the top AI stock that will likely be an integral part of AI infrastructure for years to come.
While some parts of the technology sector will likely continue to be volatile, these tech stocks could be great additions to any portfolio if you're looking to ride the long-term benefits of artificial intelligence.
Should you buy stock in Broadcom right now?
Before you buy stock in Broadcom, 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 Broadcom 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 $503,268!* Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,049,793!*
Now, it’s worth noting Stock Advisor’s total average return is 898% — a market-crushing outperformance compared to 182% 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 March 28, 2026.
Chris Neiger has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Alphabet, Amazon, Meta Platforms, Micron Technology, Microsoft, and Nvidia. The Motley Fool recommends Broadcom. 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.
AI Talk Show
Four leading AI models discuss this article
"The article mistakes cyclical capex acceleration for structural AI TAM expansion, pricing in 5+ years of 50%+ growth when memory cycles and competitive custom silicon threaten a 2027 correction."
The article conflates cyclical capex euphoria with durable competitive moats. Yes, NVDA's 86% AI chip share is real, but the article omits: (1) AMD's MI300X gaining traction with hyperscalers seeking vendor diversification; (2) custom silicon efforts at Meta, Google, and Amazon that could cannibalize 5-15% of NVDA's TAM within 24-36 months; (3) MU's $200B capex bet assumes memory demand stays parabolic—historically, memory cycles crash hard when supply normalizes. AVGO at 60x P/E pricing in perfection; one miss on ASIC adoption or hyperscaler inventory correction could trigger 30-40% drawdown. The article treats these as 'long-term holds' but ignores near-term valuation risk and cyclical headwinds.
If custom silicon adoption accelerates faster than consensus expects, or if a macro slowdown crushes capex in 2026-27, NVDA's forward multiples compress sharply despite still-strong absolute earnings—and AVGO and MU face inventory destocking that could halve revenue growth for 2-3 quarters.
"The article ignores the historical cyclicality of the semiconductor industry and the growing trend of hyperscalers insourcing chip design to bypass merchant silicon providers."
The article presents a highly optimistic view of the AI semiconductor cycle, but it glosses over significant valuation and cyclicality risks. While Broadcom (AVGO) and Nvidia (NVDA) dominate their respective niches, the 'niche' ASIC market is becoming a battleground as hyperscalers like Amazon and Google increasingly design their own silicon to reduce dependency. Furthermore, the article cites Micron (MU) as a bargain at 20x P/E, but memory is notoriously cyclical; buying at peak earnings often leads to 'value traps' when supply eventually outstrips demand. The projected $1 trillion revenue for Nvidia by 2027 assumes a linear growth trajectory that ignores potential digestion periods in capital expenditures.
If the 'AI sovereign' and enterprise demand waves are truly in their infancy, current valuations may actually be conservative as we shift from a general-purpose to an accelerated computing paradigm. A lack of viable competitors to Nvidia's CUDA software moat could sustain these margins far longer than historical hardware cycles suggest.
"N/A"
[Unavailable]
"Extravagant 2026 revenue projections and overlooked risks like capex fatigue and competition render these stocks overvalued for long-term portfolios today."
The article pitches AVGO, NVDA, and MU as long-term AI winners, flaunting explosive growth like Broadcom's AI revenue doubling to $8.4B in Q1 2026 and Micron's near-tripling to $23.9B in Q2. But it glosses over sky-high valuations—AVGO's 60x trailing P/E screams overbought (tech avg ~37x), NVDA's 35x post-570% surge assumes endless dominance, and MU's 'bargain' 20x ignores memory's boom-bust cycles. Key omissions: hyperscaler capex ($650B) exposed to ROI scrutiny, energy bottlenecks for data centers, and ASIC competition eroding Broadcom's 60% share claim. These aren't 'buys now' at current prices; they're bets on perfection in a hype-fueled sector.
If AI training/inference demand sustains hyperscaler spending ramps and Nvidia's GPU moat holds, these stocks could justify premiums with trillion-scale revenues by 2027 as projected.
"Energy infrastructure, not chip competition, is the binding constraint on hyperscaler AI capex through 2026."
Claude and Gemini both flag custom silicon cannibalization, but underweight NVIDIA's CUDA moat—software switching costs are vastly higher than hardware substitution. Grok's energy bottleneck point is concrete; data center power constraints could actually *limit* hyperscaler capex growth more than demand saturation. Nobody's quantified this. If PUE (power usage effectiveness) degradation forces capex delays, the 2027 revenue projections crater regardless of chip demand.
"Energy constraints may actually strengthen NVIDIA's moat by forcing hyperscalers to prioritize high-efficiency Blackwell chips over less power-efficient alternatives."
Claude and Grok mention energy bottlenecks, but ignore the 'Power-to-Chip' arbitrage. If power is the constraint, it actually favors NVDA's Blackwell architecture because its superior performance-per-watt allows hyperscalers to extract more compute from limited grid capacity. This makes NVDA a defensive play against energy scarcity, not a victim. Conversely, AVGO’s ASIC margins are vulnerable if power-starved data centers prioritize generic GPU density over specialized networking overhead.
[Unavailable]
"TSMC CoWoS capacity constraints will limit NVDA GPU shipments to ~2M/year vs. 5M+ demand, exacerbating energy and capex risks."
Gemini's perf/watt defense of NVDA ignores TSMC's CoWoS packaging bottleneck—capacity limited to ~2M advanced GPU equivalents/year through 2025 vs. hyperscaler demand for 5M+. This forces allocation rationing, delays Blackwell ramps, and risks parallel inventory builds like 2022. Energy constraints amplify it: power-limited DCs prioritize proven H100s over unshipped Blackwells. Supply trumps architecture here, capping the cycle sooner.
Panel Verdict
No ConsensusThe panel consensus is bearish, warning about cyclical headwinds, high valuations, and competition from custom silicon and ASICs in the AI semiconductor sector. They also highlight energy bottlenecks as a significant risk factor.
None explicitly stated
Energy bottlenecks and competition from custom silicon and ASICs