Memory has quietly become the star of the semiconductor story. The chips that store and feed data to AI accelerators: DRAM, NAND and especially high-bandwidth memory (HBM) are in such short supply that the memory segment is expected to grow roughly 30% in 2026, with leaders warning the crunch could last beyond 2030. Case in point: HBM frontrunner SK Hynix just listed ADRs on NASDAQ to ride the frenzy.
The pull is structural. Modern AI accelerators are only as fast as the memory feeding them, and the scale of this year's data-center buildouts caught the market off guard, producing a capacity crunch that some expect to persist for years. Suppliers are racing to keep up, locking in long-term wafer commitments and pouring capital into new fabs.
For investors, memory is only one layer of a fast-rotating value chain, and no single stock captures it which is where index-based exposure comes in. The MVIS US Listed Semiconductor 25 Index (MVSMH) spans the 25 largest US-listed chipmakers, including memory heavyweight Micron. The MarketVector US Listed Fabless Semiconductor Index (MVSMHX) zeroes in on the designers driving demand for all that memory. And the MarketVector China Semiconductor 25 Index (MVSMHC) opens a window into China's parallel build-out, where domestic memory and equipment makers race toward self-sufficiency.
But the theme's clarity hasn't meant a smooth ride. Over the past year the broad US semiconductor benchmark climbed sharply, only to give back a chunk of those gains this summer as memory names briefly slipped into bear-market territory before rebounding. The China index, launched only in late June 2026, has been especially volatile out of the gate. Leadership rotates constantly across designers, equipment makers, memory and foundries, a reminder that a powerful narrative still demands disciplined, diversified access.
Three lenses, one theme: as AI's appetite grows, memory and the chips around it stay foundational.
#Semiconductors #Memory #AIInfrastructure #DRAM #NAND #HBM #Investing
About the Author(s):
Kinjal Shukla is Index Researcher and Data Engineer at MarketVector IndexesTM. She is responsible for the design and development of MarketVector indexes and for creating data infographics. Kinjal has cleared FRM Level 1 having come from a risk profile in Barclays, India. She has a degree in a Master of Science in Financial Engineering from Stevens Institute of Technology, USA, and an MBA-Tech degree from NMIMS-MPSTME University, India.
For informational and advertising purposes only. The views and opinions expressed are those of the authors but not necessarily those of MarketVector Indexes GmbH. Opinions are current as of the publication date and are subject to change with market conditions. Certain statements contained herein may constitute projections, forecasts, and other forward-looking statements that do not reflect actual results. It is not possible to invest directly in an index. Exposure to an asset class represented by an index is available through investable instruments based on that index. MarketVector Indexes GmbH does not sponsor, endorse, sell, promote, or manage any investment fund or other investment vehicle that is offered by third parties and that seeks to provide an investment return based on the performance of any index. The inclusion of a security within an index is not a recommendation by MarketVector Indexes GmbH to buy, sell, or hold such security, nor is it considered to be investment advice.
Get the latest news & insights from MarketVector
Get the newsletterRelated:
.png)
