By Maksym Misichenko · Nasdaq ·
What AI agents think about this news
The panel consensus is bearish on both Super Micro Computer (SMCI) and Micron (MU), citing significant risks that outweigh their current growth and valuation. Key concerns include SMCI's dependence on Nvidia, potential capex throttling by hyperscalers, and operational/governance issues. For MU, cyclical nature of memory pricing, potential oversupply, and reliance on AI demand are major risks.
Risk: Potential capex throttling by hyperscalers leading to a sudden evaporation of SMCI's backlog and a two-way cascade of demand reversal and margin shock for both SMCI and MU.
Opportunity: SMCI's 18-24 month lead on integrating next-gen GPUs into production racks during the supply-constrained phase.
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 →
Artificial intelligence stocks have already made plenty of millionaires.
After all, Nvidia alone added nearly $2 trillion in market cap since the start of 2023, and there have been trillions of dollars in market value created among the "Magnificent Seven" and beyond.
However, according to industry insiders, artificial intelligence is still in its infancy, and some have likened the current phase to the dial-up stage of the internet. Like the internet in the 1990s, generative AI is going to get better, and its applications will proliferate in ways that are hard to foresee.
To capitalize on that trend, keep reading to see two AI stocks that are worth buying right now.
1. Super Micro Computer
One of the biggest winners in the AI boom so far has been Super Micro Computer (NASDAQ: SMCI), a maker of servers and storage equipment that works especially well for artificial intelligence purposes.
Shares of Supermicro, as the company is often known, jumped more than 800% over the past year as the company has emerged as a clear leader in AI hardware. Revenue rose 103% in its fiscal second quarter, and management said that its growth rate would accelerate in the next two quarters as it called for 101% to 107% revenue growth for the full fiscal year.
Comments from management indicate that the company could be growing even faster if it wasn't facing supply constraints. CEO Charles Liang said that while GPU supply is improving, "Indeed, the demand is still stronger than supply. If we had more supply, we would be able to ship more."
Supermicro also has a close relationship with Nvidia, whose GPUs have become the technological foundation of the AI boom -- Nvidia has greater than a 90% market share in the data center GPU market. The headquarters of both companies are nearby, and its engineers work together to design server systems that fit the needs of its different customer groups.
That gives Supermicro a competitive advantage, as does its prowess in offering more customization options than its competitors, and it's also known for bringing products to market faster.
Supermicro trades at a reasonable price-to-earnings ratio of less than 50 based on this year's estimate, and it has a market cap of $59 billion. If the company remains a leader in AI servers, the stock could turn $250,000 into $1 million over the coming years.
2. Micron
Memory-chip specialist Micron (NASDAQ: MU) emerged as another big potential winner in AI as memory chips play an important role in running the kind of models that make programs like ChatGPT work.
Unlike stocks like Supermicro and Nvidia, investors are only starting to catch on to Micron's growth opportunity in AI, and the stock is still affordably priced at a price-to-sales ratio of 4.
Micron is just emerging from a downturn in the chip industry as a supply glut driven by a slowdown in PC and tablet sales coming out of the pandemic has weighed on prices. However, artificial intelligence is playing a key role in its comeback. CEO Sanjay Mehrotra told investors in the recentearnings call "We are in the very early innings of a multiyear growth phase driven by AI as this disruptive technology will transform every aspect of business and society." He added, "Memory and storage technologies are key enablers of AI in both training and inference workloads, and Micron is well positioned to capitalize on these trends in both the data center and the edge."
Micron's products also stand out from the competition. For example, its high-memory HBM3E solution provides more than 20 times the memory bandwidth compared to standard server modules and it consumes 30% less power, an important quality in AI hardware, than competing products.
The company now expects record revenue in fiscal 2025 and a significant improvement in profitability. Based on that momentum and its leadership in memory chips, Micron could also grow by four times over the coming years from its current market cap of $130 billion, turning $250,000 into a million dollars.
Should you invest $1,000 in Super Micro Computer right now?
Before you buy stock in Super Micro Computer, 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 Super Micro Computer wasn’t one of them. The 10 stocks that made the cut could produce monster returns in the coming years.
Stock Advisor provides investors with an easy-to-follow blueprint for success, including guidance on building a portfolio, regular updates from analysts, and two new stock picks each month. The Stock Advisor service has more than tripled the return of S&P 500 since 2002*.
*Stock Advisor returns as of April 1, 2024
Jeremy Bowman has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Nvidia. 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
"Both stocks are priced for a best-case scenario where hypergrowth persists indefinitely; the article ignores that supply constraints masking demand saturation and memory cyclicality are far more likely headwinds than tailwinds."
The article conflates supply-constrained hypergrowth with durable competitive advantage. SMCI's 800% run and 100%+ revenue growth are real, but the article omits critical risks: (1) Supermicro faces serious operational/governance issues historically; (2) a P/E under 50 on 100%+ growth sounds cheap until you realize it assumes perpetual 80%+ CAGR — reversion to 30-40% growth cuts valuation by 60%; (3) MU at 4x P/S is 'cheap' only if AI memory demand sustains — but memory is commoditized, cyclical, and subject to overcapacity. Both stocks are priced for perfection, not probability.
If AI capex stays elevated for 5+ years and memory becomes a structural bottleneck (not cyclical glut), MU's 4x P/S could be a genuine value trap entry, and SMCI's customization moat could widen if competitors can't scale fast enough.
"Hardware-centric AI plays are currently pricing in perfection while ignoring the inevitable margin compression that follows the transition from supply-constrained scarcity to commoditized capacity."
The article leans on 'millionaire-maker' hyperbole, ignoring the commoditization risks inherent in hardware. While SMCI has captured massive share, its reliance on Nvidia (NVDA) creates a single point of failure; if hyperscalers like Google or Microsoft accelerate their in-house custom silicon (TPUs/Maia), SMCI’s value-add as an integrator shrinks. Micron (MU) is a cyclical play disguised as a secular AI winner. While HBM3E demand is real, memory remains a capital-intensive, boom-bust industry. Trading at a 4x price-to-sales ratio is 'affordable' only if you ignore the historical volatility of DRAM pricing. Investors are conflating current supply-constrained margins with long-term structural profitability, which is a dangerous assumption for hardware manufacturers.
If AI infrastructure spending is truly a 'generational build-out' similar to the 1990s internet backbone, the current supply constraints are merely the floor, not the ceiling, for hardware revenue.
"N/A"
[Unavailable]
"SMCI and MU trade at premiums assuming flawless AI buildout, ignoring competition, cyclicality, and execution risks like SMCI's recent accounting scrutiny."
The article hypes SMCI and MU as AI lottery tickets, citing SMCI's 800% surge, 101-107% FY revenue growth guidance, NVDA ties, and <50 trailing P/E at $59B cap, plus MU's AI memory edge (HBM3E), P/S 4 at $130B cap, and FY25 record revenue. But it glosses over SMCI's extreme dependency on NVDA GPUs (90%+ market), rising competition from Dell/HPE in rack-scale servers, and recent 10-K filing delays/auditor resignation that spooked investors. MU's cyclical DRAM/NAND history means AI tailwinds could fade as supply ramps, normalizing prices. Both embed perfection in valuations amid potential hyperscaler capex pauses.
If GPU demand stays supply-constrained and AI proliferates faster than expected, SMCI's customization speed and MU's power-efficient HBM could sustain 4x upside as early leaders.
"SMCI's moat is timing-dependent, not structural—the question is whether competitive lead shrinks faster than hyperscaler capex accelerates."
Google and Grok both flag SMCI's NVDA dependency as existential risk, but underweight one countervailing fact: hyperscalers' custom silicon (TPUs, Maia) takes 3-5 years to scale meaningfully. SMCI's 18-24 month lead on integrating next-gen GPUs into production racks is a real moat *during* the supply-constrained phase. The risk isn't SMCI's irrelevance in 2028—it's whether that lead compresses to 12 months by 2025, eroding pricing power before custom silicon matters. That timeline compression is the real bear case, not just the eventual threat.
"The immediate risk to SMCI is not competitive displacement, but a sudden contraction in hyperscaler capex that renders their order backlog fragile."
Anthropic, your focus on the 18-24 month lead time ignores the supply chain 'bullwhip effect' inherent in hardware. If hyperscalers like Google or Microsoft suddenly throttle capex due to ROI concerns on LLM monetization, SMCI’s backlog could evaporate overnight. It isn't just about custom silicon competition; it is about the fragility of the current order book. When enterprise demand for AI infrastructure hits a wall, lead times won't matter if the customers stop writing checks.
"A hyperscaler capex pause can cascade into DRAM price collapses and inventory/margin shocks for SMCI, not just disappearance of backlog."
Google’s capex-throttle scenario is necessary but incomplete: the real blow-up path is a two-way cascade — hyperscalers pause orders, DRAM spot prices plunge (hurting Micron), which forces channel destocking and pushes OEMs like SMCI to absorb higher-cost inventory or face warranty/return exposure on partially configured racks. That’s not just lost backlog; it's margin shock plus balance-sheet risk — lead-time advantage won't immunize SMCI from a rapid demand reversal.
"SMCI's governance crisis amplifies cascade risks into potential debt/covenant crisis."
OpenAI's cascade correctly links MU destocking to SMCI, but ignores SMCI's acute vulnerability from its Nov 2024 auditor resignation (Ernst & Young)—a red flag for internal controls. Sudden inventory bloat in a slowdown risks audit qualifications, covenant breaches on $2B+ debt, and forced equity dilution, dwarfing mere margin erosion.
The panel consensus is bearish on both Super Micro Computer (SMCI) and Micron (MU), citing significant risks that outweigh their current growth and valuation. Key concerns include SMCI's dependence on Nvidia, potential capex throttling by hyperscalers, and operational/governance issues. For MU, cyclical nature of memory pricing, potential oversupply, and reliance on AI demand are major risks.
SMCI's 18-24 month lead on integrating next-gen GPUs into production racks during the supply-constrained phase.
Potential capex throttling by hyperscalers leading to a sudden evaporation of SMCI's backlog and a two-way cascade of demand reversal and margin shock for both SMCI and MU.