June 25, 2026, (Inside AI) — Semiconductor stocks exploded in after-hours trading Wednesday, adding more than $400 billion in market value. The surge followed blockbuster forecasts from Micron Technology and Qualcomm, reigniting an AI rally that had recently shown signs of fatigue.
Micron projected quarterly earnings well above analyst expectations, a direct signal that massive AI infrastructure spending is translating into soaring demand for memory chips. Qualcomm, meanwhile, announced it expects $15 billion in data center chip sales by 2029, marking an aggressive pivot beyond its smartphone stronghold.
The dual announcements reversed a sharp selloff that had erased 8% from the PHLX Semiconductor Index just a day earlier. That drop reflected mounting anxiety that AI valuations had become unsustainable and that the payoff from data center investments remained distant.
Memory Makers and AI Architects Ride the Wave
Micron’s after-hours jump of 12% pulled competitors along with it. Western Digital, Sandisk, and Seagate Technology each surged more than 8%. The rally extended to chip designers and equipment suppliers: Arm Holdings rose about 6%, Marvell Technology added nearly 4%, and Broadcom climbed 2%.
Equipment giants Applied Materials and ASML both gained over 4%, underscoring how the AI boom is lifting the entire semiconductor supply chain. The moves added hundreds of billions in market capitalization within hours, a stark contrast to Tuesday’s gloom.
Qualcomm’s data center target represents a fundamental shift. The company, long synonymous with mobile processors, is now betting that its AI expertise can capture a slice of the cloud infrastructure market dominated by Nvidia and AMD. The $15 billion figure, while aspirational, signals confidence that AI workloads will demand diverse silicon solutions.
Valuation Jitters Meet Unrelenting Demand
The rally confronts a persistent tension. Even after this week’s volatility, the PHLX chip index remains up 90% in 2026. Micron alone has soared over 260% this year, not counting Wednesday’s late surge. Such gains invite scrutiny about whether stock prices have outrun fundamentals.
Yet the forecasts from Micron and Qualcomm provide concrete evidence that AI spending is not merely a speculative bubble. Memory chips are the literal foundation of AI training and inference, and Micron’s outlook suggests that hyperscalers and enterprises are still in the early stages of a multiyear buildout.
Qualcomm’s move also highlights a broader industry trend: the blurring of lines between mobile, edge, and data center computing. As AI models become more efficient and inference shifts to the edge, Qualcomm’s expertise in low-power, high-performance chips could become a strategic advantage.
The after-hours fireworks serve as a reminder that the AI trade, while prone to sharp corrections, is underpinned by tangible demand signals. For investors, the question is no longer whether AI will transform computing, but how quickly—and which companies will capture the value.
