Memory chip stocks have emerged as the next big wave of the AI trade and investors are all in.<\/p>\n
\n Loading audio narration…\n <\/p>\n
AI is driving up demand for memory, sending the stocks of the companies that make memory chips soaring and helping to lead the S&P 500 and Nasdaq Composite to fresh records.<\/p>\n
The Roundhill Memory ETF, which trades under the ticker, DRAM, demonstrated the group’s meteoric rise. DRAM has gained around 88% since it began trading just over one month ago on April 2.<\/p>\n
Roundhill CEO, Dave Mazza, told Business Insider that the firm created the fund because there was a gap in the market for a memory-focused ETF.<\/p>\n
“Memory chips are the most supply-constrained layer of the AI infrastructure buildout, but investors had no efficient way to access the theme,” he explained.<\/p>\n
The ETF is made up of several global memory chip company stocks including US names like Sandisk<\/a> and Micron<\/a> along with Korean companies like SK hynix<\/a> and Samsung<\/a>, Japan’s Kioxia Holdings<\/a>, as well as Taiwan-based Nanya Technology and Winbond Electronics.<\/p>\n “The heartbeat of AI”<\/p>\n The S&P 500 and the Nasdaq Composite are trading at record highs, largely shrugging off war in Iran<\/a> and elevated oil prices<\/a>\u2014 and memory stocks are leading the charge, seeing some of the largest year-to-date gains in the index.<\/p>\n Sandisk has been the top S&P 500 performer so far in 2026, gaining a whopping 558%.<\/p>\n Seagate<\/a> has risen 172% while Western Digital<\/a> and Micron more than doubled, gaining 156% and 137%, respectively.<\/p>\n AI is evolving from training to inference<\/a>, fueling a surge in demand for memory as the tech relies on data which is stored on memory chips, especially in its inference stage.<\/p>\n Ruben Dalfovo, an investment strategist at Saxo, said<\/a> this “shift makes memory more strategic, because inference needs speed, bandwidth and power efficiency, not just raw compute.”<\/p>\n Micron explained<\/a> that memory and storage components are “the\u202fheartbeat of AI,” describing the components as “serving as its cognitive backbone, enabling innovation, accelerating performance, and redefining their traditional role from commodity to strategic asset.”<\/p>\n There’s not enough memory to go around<\/p>\n AI is driven memory demand sky high to the point where supply can’t keep up.<\/p>\n “The memory market is at an unprecedented inflexion point, with demand materially outpacing supply,” IDC, a tech market intelligence firm, said.<\/a><\/p>\n During the AI boom, memory chipmakers have prioritized making high bandwidth memory solutions over making consumer electronic components, lifting prices across the board.<\/p>\n “AI servers and enterprise environments require far more memory per system than consumer devices, so the AI build-out is pulling a disproportionate share of global capacity and creating shortages, as suppliers prioritize orders from hyperscalers and OEMs building AI servers,” IDC explained.<\/p>\n The firm cautioned that this memory cycle is different than previous tech waves as AI is reshaping the market itself. “This is not just a cyclical shortage driven by a mismatch in supply and demand, but a potentially permanent, strategic reallocation of the world’s silicon wafer capacity,” they said.<\/p>\n “Memory chips are the AI bottleneck,” Roundhill’s Dave Mazza told Business Insider.<\/p>\n “Demand from hyperscalers is non-discretionary, supply is physically constrained with new fabs taking three to five years to come online, and Micron’s entire 2026 HBM allocation is sold out under fixed pricing agreements,” he added.<\/p>\n The supply-demand imbalance sends memory prices soaring<\/p>\n While demand outpaces supply, memory chip prices have surged.<\/p>\n In fact, memory cost headwinds were a trend in Big Tech earnings<\/a>.<\/p>\n Meta CEO Mark Zuckerberg said the jump in the company’s spending<\/a>, was due in part to higher memory chip costs. Apple also called out high memory costs as a headwind.<\/p>\n