AI Memory Demand Drives Micron and Sandisk as Wall Street Rethinks the Cycle

AI infrastructure is reshaping the outlook for memory chips

The memory trade regained momentum as investors focused on stronger pricing and growing demand from artificial-intelligence infrastructure. Micron Technology (MU) closed Friday up 2.3% at $970.20 for its fourth consecutive gain, while Sandisk (SNDK) surged 7.39% following an investor day that outlined ambitious long-term growth and margin targets.

The moves reflect a broader shift taking place across memory markets. KeyBanc expects DRAM prices to rise 15% to 20% in the third quarter and another 15% in the fourth quarter, while NAND flash prices are projected to increase 30% to 40% in the third quarter followed by another 15% gain in the fourth. At the same time, AI workloads are consuming growing amounts of high-performance memory and data-center storage.

Key Points

• Micron gained for a fourth consecutive session as tight supply and AI demand supported expectations for continued increases in DRAM and NAND pricing.
• Sandisk’s investor day highlighted a rapidly expanding NAND market, with the company targeting mid-to-high teens revenue growth and adjusted gross margins around 80% between 2028 and 2030.
• Wall Street is increasingly questioning whether AI, long-term customer agreements and high-bandwidth memory could reduce the severity of the traditional memory boom-and-bust cycle.

Rising Memory Prices Bring Investors Back to Micron and Sandisk

Memory stocks strengthened as expectations for higher pricing reinforced the industry's near-term outlook.

KeyBanc forecasts DRAM prices will increase 15% to 20% sequentially during the third quarter and another 15% in the fourth. NAND flash prices are expected to rise even faster, increasing between 30% and 40% in the third quarter before gaining another 15% in the fourth.

Those projections come as AI data-center construction consumes large amounts of memory capacity. Much of Micron's 2027 capacity has already been allocated as demand continues to exceed supply.

Micron's latest fiscal third quarter demonstrated the financial impact. Revenue reached $41.46 billion, compared with $23.86 billion in the previous quarter, representing roughly 74% sequential growth. Adjusted free cash flow reached $18.3 billion after $7.1 billion of capital spending, while cash, investments and restricted cash totaled $30.2 billion.

Data-center revenue exceeded $25 billion during the quarter, representing an annualized rate above $100 billion, while HBM4 shipments surpassed $1 billion. Adjusted gross margin reached 84.9%, with operating margin at 81.2%.

Micron has also established approximately $100 billion of contracted revenue across 16 Strategic Customer Agreements. Those agreements cover roughly 20% of DRAM volume and one-third of NAND volume.

Sandisk's latest results highlighted a parallel acceleration in NAND.

Revenue reached $8.965 billion, with approximately one-third of the increase coming from volume and two-thirds from pricing. Data-center revenue increased 103% sequentially.

Sandisk has approximately $93.9 billion of minimum contracted revenue across 10 agreements, supported by $16.5 billion in financial guarantees.

The combination of rising prices, contracted demand and AI infrastructure spending is giving investors additional evidence that current memory conditions differ from previous periods driven primarily by consumer electronics.

Why Is AI Changing the Traditional Memory Cycle?

Memory has historically been one of the semiconductor industry's most cyclical businesses.

When demand increased, manufacturers expanded production. Eventually supply exceeded demand, prices declined and earnings contracted. Because investors expected strong profits to disappear during the next downturn, memory companies typically received relatively low valuation multiples.

AI is challenging parts of that framework.

High-bandwidth memory, or HBM, has become an important component of AI accelerators. It is more technologically complex than conventional DRAM, requires advanced packaging and consumes substantially more wafer capacity.

HBM uses nearly three times the wafer capacity per gigabyte compared with conventional DDR5 DRAM because it requires larger dies and multiple vertically stacked DRAM layers.

That creates an unusual supply dynamic. Rising HBM production satisfies growing AI demand while simultaneously consuming manufacturing capacity that otherwise could be used to produce conventional memory.

New Street Research analyst Pierre Ferragu argues this could make HBM structurally less cyclical than commodity DRAM.

His firm models AI eventually accounting for nearly two-thirds of memory demand and expects overall memory demand to grow approximately 15% annually beyond 2030, compared with roughly 10% historically.

New Street upgraded Micron to Buy from Neutral and raised its price target to $1,250. The firm estimates Micron could hold more than $600 billion in cash by 2030 while producing more than $150 billion in annual free cash flow at peak levels.

Even during a modeled four-year downturn beyond 2030, New Street expects a trough of only $18 billion in free cash flow burn and more than $100 billion of annual generation during the period.

UBS is also looking beyond near-term peak earnings. Analyst Timothy Arcuri set a $1,625 Micron price target based on 11 times his forecast for 2029 earnings, a period his model assumes includes a moderate memory downturn.

The argument is not that memory cyclicality disappears entirely. Instead, analysts are examining whether AI demand, capacity constraints and longer-term customer agreements could reduce the amplitude of future cycles.

Sandisk Signals a Structural Shift in NAND Demand

Sandisk's investor day added another dimension to that argument by focusing on the transformation taking place in NAND flash memory.

The company projects its market could reach $300 billion this year and approach $500 billion in 2027, compared with approximately $70 billion in 2025.

For 2028 through 2030, Sandisk is targeting mid-to-high teens revenue growth, adjusted gross margins of roughly 80% and adjusted operating margins of 75%.

J.P. Morgan resumed coverage with an Overweight rating and a $2,250 price target, while Citi reiterated a Buy rating with a $2,100 target.

A central part of the thesis is the growing importance of AI inference.

Inference occurs when trained AI models use what they have learned to process information and generate results. As models handle increasingly large amounts of context, they create significant memory requirements.

One emerging application involves key-value, or KV, caching. AI models generate key-value pairs while processing tokens, allowing them to reuse previous context rather than repeatedly performing the same computational work.

Those caches traditionally reside in expensive HBM or DRAM. As context windows and workloads grow, however, data centers can move portions of the cache into NAND flash storage, where information can remain for extended periods at lower cost.

That gives NAND an expanding role inside AI infrastructure rather than leaving it primarily dependent on traditional consumer storage markets.

Sandisk is simultaneously shifting from shorter-term sales toward long-term guaranteed customer contracts, which could increase financial predictability.

The company's investor-day outlook helped lift other memory and storage stocks as investors reconsidered the potential duration of the current demand environment.

Micron rose alongside Sandisk, while Western Digital (WDC) and Seagate Technology Holdings (STX) also advanced.

What It Means for Investors

The central issue surrounding Micron and Sandisk is becoming less about whether memory demand is currently strong and more about whether AI has structurally changed the industry's economics.

There is substantial evidence of tight conditions. DRAM and NAND prices are expected to continue rising through the second half of the year, Micron has already allocated much of its 2027 capacity, and both companies are establishing substantial long-term customer commitments.

The composition of demand is also changing.

Micron has direct exposure to HBM and DRAM used in AI accelerators, while Sandisk is benefiting from expanding NAND requirements across data centers and inference workloads. Both are increasingly tied to infrastructure spending rather than relying exclusively on traditional consumer electronics cycles.

Competition remains important.

Micron slipped behind Chinese rival YMTC in global NAND shipment share during the second quarter. Samsung led with 25%, followed by SK Hynix at 22% and YMTC at 14%, with Micron trailing.

Shipment volume, however, did not translate directly into revenue leadership. YMTC ranked fifth in NAND revenue because its product mix remains concentrated in lower-priced consumer applications and has less exposure to higher-priced data-center enterprise solid-state drives.

Valuation also remains complicated because memory earnings can change dramatically across cycles. Micron trades around 13 times forward non-GAAP earnings according to one valuation measure, while another calculation places it at 12.4 times current-year earnings and 5.9 times projected fiscal 2027 earnings.

The usefulness of those multiples depends heavily on whether today's elevated earnings prove temporary or represent a higher level of through-cycle profitability.

That is why Wall Street's changing view of memory cyclicality matters. If AI demand, HBM capacity requirements and long-term contracts make earnings more durable, traditional peak-cycle valuation frameworks may become less representative. If supply eventually overwhelms demand again, the industry's familiar cyclicality remains relevant.

Conclusion

Micron and Sandisk are benefiting from the same broad force—rapidly expanding AI infrastructure—but through different parts of the memory market.

Micron's story centers on HBM and DRAM, where AI accelerators are consuming large amounts of high-performance memory and limiting available manufacturing capacity. Sandisk's opportunity is concentrated in NAND, where data centers and inference workloads are expanding the role of flash storage.

Both companies are also using long-term customer agreements to increase revenue visibility in an industry historically defined by short-term pricing swings.

The immediate market signal remains strong. DRAM and NAND prices are expected to rise further, Micron's capacity remains constrained, and Sandisk's long-term outlook points toward substantial expansion in AI-related storage demand.

The longer-term question is whether those changes are powerful enough to reshape memory's traditional economics.

That distinction will determine whether the current environment ultimately proves to be an unusually large memory upcycle or something more structural.

FAQs

Why did Micron stock rise?

Micron shares rose as investors responded to expectations for continued increases in memory pricing, strong AI-related demand and Sandisk's upbeat long-term outlook for the memory market. Micron closed Friday up 2.3% at $970.20 for its fourth consecutive gain.

What are the forecasts for DRAM and NAND prices?

KeyBanc expects DRAM prices to increase 15% to 20% in the third quarter and another 15% in the fourth quarter. NAND prices are projected to rise 30% to 40% in the third quarter followed by another 15% increase in the fourth.

How is AI changing Micron's memory business?

AI is increasing demand for high-bandwidth memory used in accelerators and data centers. HBM also consumes substantially more wafer capacity than conventional DRAM, potentially limiting overall memory supply while AI demand continues to grow.

Why did Sandisk stock rise?

Sandisk shares climbed after the company outlined long-term targets including mid-to-high teens revenue growth between 2028 and 2030, adjusted gross margins around 80% and adjusted operating margins of 75%, supported by expanding AI data-center and inference demand.

What is the biggest question facing Micron and Sandisk?

The key question is whether AI demand, tighter capacity and long-term customer agreements can make memory earnings more durable and reduce the severity of the industry's traditional boom-and-bust cycles.

This article was created with AI assistance and reviewed by an editor. For details, please refer to our Terms of Use.

Go Beyond the Market Brief with Market Edge

Follow SharperTrades’ complete approach to trading and investing, combining active trade opportunities through Block Orders, long-term research through Stock Investor, and structured market education through the Swing Trading Masterclass. Try Market Edge for $19 your first month →

Explore Research with Stock Investor

Stock Investor is SharperTrades’ platform for long-term investing research and portfolio management. Members receive research reports, portfolio updates, conviction tracking, and in-depth analysis designed to support disciplined investment decisions.

Explore Active Trading & Income Strategies

Block Orders tracks institutional activity and highlights active trade setups and price behavior across long and short opportunities.

For options-focused traders, Essential Option Income provides a structured approach to options income strategies, while Pro Option Trader offers a broader range of options strategies and trade opportunities.

Think More Clearly with SteadyCapital

SteadyCapital is SharperTrades’ decision-support system for long-term investors, built around the SteadyCapital Method™. Review investment ideas, challenge assumptions, evaluate valuation and risk, compare companies, and think through important buy, hold, add, trim, or sell decisions before you act.

Risk Disclosure

All content is provided for educational purposes only and does not constitute investment advice. Trading involves risk, and past performance is not indicative of future results. Please review our full Risk Disclosure for additional information.