By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel consensus is overwhelmingly bearish on the article's bullish AI thesis, citing cyclical semiconductor industry trends, questionable data, and unaddressed risks such as memory cycle expansions and power constraints.
Risk: Rapid margin compression and multiple de-rating due to supply catching up faster than AI demand.
Opportunity: None identified by the panel.
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 →
Micron and Sandisk are thriving from the memory chip shortage.
CoreWeave will see monster growth over the next few years as more data centers come online.
If you've got $5,000 sitting around waiting to invest, now could be a smart time to put it to work. Several high-growth investment opportunities could easily provide solid upside in the short term, but also represent solid long-term picks if the current trend lasts over the next five years.
Three high-growth stocks that I'm eyeing are Sandisk (NASDAQ: SNDK), Micron (NASDAQ: MU), and CoreWeave (NASDAQ: CRWV). All three of these are rapidly growing and are thriving in the artificial intelligence (AI) buildout.
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 »
Micron manufactures both DRAM and NAND memory chips, each of which is in short supply. Memory chip demand is driven by the massive AI build-out, which is causing the prices on these chips to spike.
Micron is benefiting from this, but it's also building out extra manufacturing capacity to meet demand. However, those facilities won't be operational until later next year, which means the memory chip shortage could last for a few more years. This shortage gives Micron investors an opportunity to make a ton of money, as it's rapidly growing due to soaring commodity prices.
Next quarter, analysts expect 264% revenue growth. For the fourth quarter of fiscal year 2026 (ending in August), they expect an additional 250% growth. Those are solid figures, yet Micron still trades at a discount to most of its tech peers, which commonly trade for 20 to 30 times forward earnings.
Micron looks like a strong growth and value play now, and with the memory chip shortage expected to last for a few more years, it's a great pick.
Sandisk is in a similar boat as Micron, but it only makes NAND memory, which typically gets consumed in solid-state drives. Solid-state drives are important in data centers for long-term data storage and are similarly experiencing a shortage, driving prices to soar. Sandisk's revenue growth is more rapid than Micron's, with Wall Street analysts projecting 332% and 337% growth over the next two quarters.
Sandisk is also more expensive than Micron at 28 times forward earnings, but it may deserve that premium with the higher growth rate. Both Sandisk and Micron will continue to see strong growth for the foreseeable future until the memory supply increases.
However, AI hyperscalers are also spending more on data center capital expenditures each year, so just because more production capacity is being built, it doesn't mean the memory chip shortage will be resolved anytime soon. That could make both Micron and Sandisk strong multiyear plays, which is why I think they are both solid stock picks now.
Switching gears a bit, CoreWeave is one of the companies causing the memory chip shortage. It operates several data centers and fills them with cutting-edge GPUs, and rents out the computing capacity to its clients. It's seeing strong demand for its cloud computing products, and has captured several major clients, like Meta Platforms and Microsoft.
CoreWeave is also seeing strong growth, and Wall Street estimates that its next two quarters of growth will be 112% and 154%. This strength will likely last for several more years, as CoreWeave has a gigantic backlog to churn through.
It has nearly $100 billion in revenue contracted over about a five- to six-year time frame, and that figure will likely expand with each quarter as new capacity comes online and new clients are onboarded. That will lead to phenomenal growth for CoreWeave over the next few years. As long as the AI buildout continues to gain momentum, CoreWeave will be an excellent investment, as it's working to build a computing footprint as large as possible before the AI arms race is over. If it can capture several major clients, it will have a long-term, continuous revenue stream that will make it a top AI stock to own.
Before you buy stock in Micron Technology, 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 Micron Technology 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 $443,191! Or when Nvidia made this list on April 15, 2005... if you invested $1,000 at the time of our recommendation, you’d have $1,258,838!
Now, it’s worth noting Stock Advisor’s total average return is 941% — a market-crushing outperformance compared to 206% 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 June 7, 2026. *
Keithen Drury has positions in Meta Platforms and Microsoft. The Motley Fool has positions in and recommends Meta Platforms, Micron Technology, and Microsoft. 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
"The article overstates durability of the memory shortage, relies on dubious company metrics (e.g., CRWV, 100B backlog), and risks a sharp reversion in memory pricing that could crater the proposed upside."
The piece fuses a bullish AI thesis with questionable data and naming inaccuracies. It hinges on a prolonged memory-chip shortage lifting MU and SNDK, and on CoreWeave's public-market status and $100B backlog — both claims that clash with typical industry cycles and known company footprints. Memory demand is episodic; as capex cycles turn, pricing and margins compress. Even if AI compute spend remains robust, many of these names face elevated valuation, dilution risk, and competitive pressure from Samsung/SK Hynix and hyperscalers. The CRWV ticker and Sandisk branding cast further doubt on reliability; due diligence is essential before buying.
If AI demand accelerates beyond current expectations and these firms win meaningful market share, the growth deltas could still materialize despite the red flags.
"The article's investment thesis is built on outdated tickers and inaccessible private equity, while ignoring the inherent cyclicality of the memory chip market."
This article is fundamentally flawed. Most glaringly, it lists Sandisk as a public stock (SNDK), which was acquired by Western Digital in 2016. Furthermore, CoreWeave is a private company, making it inaccessible to the average retail investor looking to deploy $5,000. While Micron (MU) is a legitimate play on High Bandwidth Memory (HBM) demand, the article ignores the cyclical nature of the semiconductor industry. Memory is a commodity; when supply finally catches up to the AI-driven demand surge, margins will compress rapidly. Relying on 200%+ revenue growth projections in a cyclical, capital-intensive sector is dangerous for long-term investors.
If the 'AI supercycle' creates a permanent structural shift in data center architecture, memory makers could move from cyclical commodities to essential, high-margin infrastructure providers.
"The article treats a cyclical shortage-driven margin expansion as a secular AI growth story, ignoring that new capacity is already coming online and hyperscalers are reducing their exposure to commodity memory pricing."
This article conflates a cyclical shortage with secular growth and cherry-picks growth rates that are mathematically unsustainable. The 264-332% YoY revenue projections for MU and SNDK are real but occur off a depressed base—they don't imply 264% growth *next year*. More critically: the article assumes memory scarcity persists 'for a few more years,' but new fab capacity (TSMC, Samsung, Intel foundry) is already ramping. CoreWeave's $100B backlog sounds massive until you realize it's 5-6 year contracted revenue (~$17B annually)—impressive but not 'monster growth' at scale. The article also ignores that hyperscalers are vertically integrating chip design (Google TPUs, Amazon Trainium) to reduce dependency on spot purchases. Valuations matter: SNDK at 28x forward P/E assumes the shortage persists; if it normalizes in 18-24 months, multiple compression could erase gains.
Memory chip cycles are notoriously hard to time, and if demand truly outpaces supply through 2027-2028 as the article suggests, these stocks could re-rate higher before any normalization hits. CoreWeave's backlog is genuinely contracted revenue, not vaporware.
"Memory pricing cycles have repeatedly turned from shortage to glut within two years once new fabs come online."
The article touts MU, SNDK, and CRWV as AI-driven growth plays on memory shortages and data-center demand, citing 250-337% revenue growth forecasts through 2026. Yet it ignores the classic semiconductor cycle: capacity expansions by multiple players historically trigger oversupply and price collapses within 18-24 months. SNDK has not traded publicly since 2016 and CRWV remains private, undermining credibility. Micron's forward multiples already embed much of the near-term upside, leaving limited margin of safety if hyperscaler capex slows or competitors like Samsung ramp faster than expected.
If AI training and inference workloads scale beyond current forecasts, the memory deficit could extend into 2027, validating the multi-year growth thesis and pushing MU earnings well above consensus.
"CRWV’s public-ticker claim is dubious, but the more decisive risk is that a memory-cycle unwind could erase the article’s backbone assumptions."
ChatGPT's critique hinges on labeling CoreWeave (CRWV) as a public ticker and Sandisk branding; that factual misstep weakens the piece and distracts from real risks. The bigger, under-addressed risk is the memory cycle itself: if capacity expands faster than AI demand, the supposed 264-332% growth and a $100B backlog collapse into compressed margins and multiple de-rating. Without credible backing, the bullish thesis remains fragile.
"Power and cooling constraints act as a physical bottleneck that will likely collapse the aggressive revenue growth projections for memory and compute providers."
Claude is correct about the hyperscaler vertical integration, but misses the secondary risk: power constraints. Even if memory supply catches up, data center scaling is hitting a hard wall in utility capacity. This ‘power bottleneck’ creates a ceiling on how much HBM and compute can actually be deployed, regardless of chip availability. This makes the article’s aggressive revenue projections fundamentally disconnected from the physical reality of grid infrastructure and cooling limitations currently stalling massive AI cluster builds.
"Power is a real constraint but solvable; valuation compression from margin normalization is the actual downside risk."
Gemini's power constraint angle is real but overstated as a *ceiling* on near-term growth. Hyperscalers are actively upgrading grid infrastructure—Microsoft/Google/Meta capex budgets explicitly include power. The bottleneck delays deployment, not eliminates it. More pressing: nobody's flagged that MU's HBM margins are already compressing as supply normalizes. If MU trades at 15x forward P/E (vs. 22x today) by late 2025, the stock underperforms despite revenue growth. Timing matters more than the cycle itself.
"Power delays accelerate hyperscaler vertical integration, shortening MU's growth window rather than extending it."
Power bottlenecks flagged by Gemini could extend shortages by slowing builds, yet this same friction pushes hyperscalers toward internal chip designs faster, as Claude referenced with TPUs and Trainium. The result is reduced external memory purchases precisely when new capacity from Samsung and others comes online. MU's projected margins then face dual pressure from both supply normalization and demand substitution, eroding the multi-year thesis before 2026 peaks.
The panel consensus is overwhelmingly bearish on the article's bullish AI thesis, citing cyclical semiconductor industry trends, questionable data, and unaddressed risks such as memory cycle expansions and power constraints.
None identified by the panel.
Rapid margin compression and multiple de-rating due to supply catching up faster than AI demand.