Artificial intelligence has helped carry major US equity indices towards record levels, but the next test of the rally may come from a component that once attracted far less attention: memory.
Micron Technology is scheduled to report its fiscal third-quarter results on Wednesday, 24 June. Investors will examine the results for evidence that demand for high-bandwidth memory, data-centre infrastructure and AI-related computing remains strong enough to support elevated semiconductor valuations.
Micron produces DRAM, NAND and specialised memory products used across data centres, computers, mobile devices, vehicles and other technology systems.
Memory was traditionally considered one of the more cyclical parts of the semiconductor industry. Prices could rise rapidly when supply became limited, then fall when manufacturers expanded production or demand weakened.
AI is changing that structure. Advanced processors require large amounts of fast memory to handle complex models and data-intensive workloads. High-bandwidth memory, commonly known as HBM, has therefore become an important part of AI infrastructure rather than a standard interchangeable component.
Micron’s results may offer information about demand from data centres and cloud providers, pricing conditions across memory products, the availability of advanced HBM capacity, capital expenditure and the sustainability of AI infrastructure spending.
Micron reported fiscal second-quarter revenue of $23.86 billion, compared with $13.64 billion in the previous quarter and $8.05 billion during the same period last year.
The company also recorded a GAAP gross margin of 74.4% and issued third-quarter revenue guidance of $33.5 billion, plus or minus $750 million. Its guidance included an expected gross margin of approximately 81%.
These figures demonstrate the strength of the current memory cycle, but they also raise the threshold for another positive market surprise. When a company’s share price and earnings expectations rise quickly, strong results may no longer be enough by themselves. Investors may focus more heavily on forward guidance, order visibility, supply constraints and whether growth can continue at the same pace.
The wider memory market is also showing how AI is changing the semiconductor industry.
SK Hynix overtook Samsung Electronics to become South Korea’s most valuable listed company on 22 June. The company has benefited from its leading position in HBM products used in advanced AI systems.
Its shares have risen more than 340% this year, while its market capitalisation reached approximately $1.35 trillion. Reuters reported that SK Hynix controlled around 61% of the global HBM market in 2025, compared with approximately 21% for Micron and 17% for Samsung.
This development illustrates a broader change. Memory chips are becoming strategically important components with higher technical barriers, closer integration with AI processors and potentially stronger pricing power.
Micron shares have gained approximately 298% this year, while the Philadelphia Semiconductor Index recently reached a record high. Major US stock indices have also remained close to their all-time highs despite volatility across technology shares.
This means Micron’s report could influence sentiment beyond memory manufacturers. Strong demand, supportive pricing and confident guidance could reinforce expectations for semiconductor suppliers, data-centre companies and other AI infrastructure businesses.
However, evidence of slowing orders, weaker pricing or higher production costs could raise questions about whether parts of the AI rally have moved ahead of underlying fundamentals. The reaction may depend less on the headline earnings figure and more on what management says about future demand.
Large technology companies continue to invest heavily in computing capacity, data centres, networking equipment and advanced chips.
Combined AI-related spending by major technology companies is expected to exceed $700 billion this year, compared with around $400 billion in 2025. This expansion has supported semiconductor demand and corporate earnings across the AI supply chain.
The central question is whether this spending is creating sustainable demand or encouraging manufacturers to expand capacity too aggressively. Chip production requires long investment periods, and supply decisions made today can influence pricing several quarters or years later. This is one reason the semiconductor industry has historically experienced strong cycles.
The AI investment story is unfolding alongside a more challenging interest-rate environment.
Higher bond yields can pressure technology valuations because a greater portion of their expected value depends on future earnings. When discount rates rise, investors may become less willing to pay elevated multiples for long-term growth.
The US Personal Consumption Expenditures inflation report is also scheduled for Thursday, 25 June. A stronger inflation reading could reinforce expectations that interest rates will remain restrictive, while a softer result may provide some relief to growth-sensitive equities.
This creates two separate tests for technology markets during the week: Micron will test the strength of AI-related corporate demand, while US inflation data will test the monetary-policy environment supporting equity valuations.
Traders will first examine Micron’s revenue, margins and earnings relative to the company’s previous guidance.
Management commentary on HBM availability, customer demand and memory pricing may be more influential than the reported numbers alone.
Capital expenditure will also matter. Higher investment can signal confidence in future demand, but aggressive expansion could eventually create concerns about excess supply.
Market participants will monitor the wider response across semiconductor manufacturers, AI infrastructure companies and major technology indices. Treasury yields and Thursday’s PCE inflation report could determine whether company-specific optimism translates into a broader equity-market move.
Micron produces memory products used alongside advanced processors in AI servers and data centres. AI systems require large amounts of fast memory to process complex workloads efficiently.
High-bandwidth memory is a specialised type of memory designed to transfer large quantities of data rapidly while using less power. It is commonly integrated with advanced processors used for AI and high-performance computing.
Markets often price expected growth into a company’s shares before results are released. If expectations are extremely high, even strong figures may disappoint investors when guidance or demand commentary falls short.
Higher interest rates increase the discount applied to future earnings. This can place greater pressure on companies whose valuations depend heavily on long-term growth expectations.
