AI Data-Center Gold Rush Sends Chipmakers to New Market Highs

Semiconductor stocks extended their lead over the broader market this week, powering the Nasdaq higher as investors bet that a historic boom in artificial‑intelligence spending will keep chip demand running hot. The PHLX Semiconductor Index recently notched its highest level since last fall and is up roughly 20% so far this year, outpacing the S&P 500 as money pours back into companies supplying the processors, memory and manufacturing tools that underpin AI data centers. The latest catalyst came from a series of bullish signals on capital spending. Microsoft has outlined plans to invest about $80 billion in AI‑enabled data centers in its 2025 fiscal year, reassuring investors that cloud giants, or “hyperscalers,” are not dialing back on infrastructure budgets. Morningstar analysts estimate that collective annual AI investment by major hyperscalers — Microsoft, Alphabet, Amazon, Meta Platforms and Oracle — could triple to about $450 billion by 2027 from $150 billion in 2023, while other industry research cited by Benchmark puts 2025 AI infrastructure spending closer to $650 billion globally. Chipmakers most exposed to that spending have been among the biggest winners. Nvidia, which dominates the market for AI accelerators used in data centers, has surged on the back of record data‑center revenue and recently crossed the $5 trillion market‑capitalization threshold, becoming the world’s most valuable company in late 2025. Broadcom, a key supplier of custom AI chips and networking silicon, is up more than 40% this year, while Qualcomm, Intel and Taiwan Semiconductor Manufacturing Co. have all rallied on optimism that AI workloads will drive demand for both cutting‑edge processors and the older‑generation nodes needed for connectivity and support chips. Memory and equipment makers are riding the same wave. Micron Technology has been one of the sector’s standout performers, with its stock more than doubling over the past year on expectations that high‑bandwidth memory used in AI servers will remain in structural shortage. Equipment vendors such as Applied Materials, Lam Research, KLA and ASML have also risen sharply — in some sessions gaining 5% or more — as investors anticipate years of elevated capital expenditure on fabs, lithography systems and advanced packaging lines required to support AI chips. Analysts argue that, for now, demand still outstrips supply. “We’re still bullish on AI and semiconductors in general, led by Nvidia, as AI demand still seems to be exceeding supply,” Morningstar senior equity analyst Brian Colello said in a recent report, adding that Oracle’s and Meta’s cloud build‑out plans have been “eye‑popping” and that there are “no signs of a slowdown in hyperscaler spending.” Nvidia Chief Executive Jensen Huang has gone further, suggesting annual data‑center spending worldwide could reach as much as $4 trillion by 2030 as enterprises re‑architect infrastructure around AI workloads. Still, the speed and scale of the rally are prompting questions about how long the cycle can run without a reset. Semiconductors remain a cyclical industry, and strategists warn that even if AI demand continues to grow, the rate of new investment could decelerate from today’s breakneck pace, potentially exposing richly valued stocks to pullbacks. For now, investors appear willing to look through those risks: chipmakers and their suppliers are again the market’s momentum leaders, and the AI data‑center boom shows few signs of slowing as companies race to deploy the computing power behind the next generation of generative AI applications.