The artificial intelligence rally that made Nvidia Corp. (NVDA  ) a market juggernaut appears to have entered a new phase in 2025, as a sharp rotation in semiconductor leadership has seen rivals outperform the AI front-runner.

Despite rising nearly 14% through the end of June, Nvidia now lags behind a group of chipmakers delivering significantly stronger gains.

Crucially, Nvidia's long-held outperformance over the broader chip sector-tracked by the Invesco PHLX Semiconductor ETF (SOXQ  )-has nearly evaporated in the first half of the year.

Is this just a pause in Nvidia's dominance, or the beginning of a more permanent shift in the AI hardware race?

Nvidia Is No Longer The Top Chip Gainer

After dominating the AI-fueled bull runs of 2023 and 2024, Nvidia has posted a modest year-to-date gain of 13.6% as of June 25, fully recovering from a Deepseek-driven selloff in January.

But while Nvidia is in the green, several rival chipmakers have surged ahead.

Micron Technology Inc. (MU  ) leads the pack, up 44.5% through June. KLA Corp. (KLAC  ) follows with a 40.4% gain, and Lam Research Corp. (LRCX  ) has jumped 32.9%.

Other notable performers include Arm Holdings PLC (ARM  ) with a 24% rise, Advanced Micro Devices Inc. (AMD  ) up 19.1% and Broadcom Inc. (AVGO  ) gaining 14.7%.

By contrast, Taiwan Semiconductor Manufacturing Co. (TSM  ) has added 13%, nearly matching Nvidia, while the broader sector tracked by the Invesco PHLX Semiconductor ETF rose 10%.

This marks a reversal from 2024, when Nvidia soared 179.3%, more than doubling the returns of TSM, which gained 93.8%.

In 2023, Nvidia delivered a staggering 236.5% return, leaving AMD (+116.5%) and Broadcom (+100.1%) in the dust. That level of dominance is now being challenged.

Why Is The AI Leadership Rotating?

The sector rotation may reflect a broadening investor focus across the AI hardware stack.

While Nvidia continues to dominate AI accelerators, analysts see expanding opportunities in memory, manufacturing, and equipment, where companies like Micron, Lam and KLA are well-positioned.

In a note shared Wednesday, Bank of America analyst Vivek Arya estimated that the global AI data center market could grow from about $250 billion today to about $1 trillion by 2030, reflecting a 26% compound annual growth rate.

That growth rate is more than three times faster than overall IT spending, which Arya expects to rise at an 8% pace.

More than $800 billion of the projected $1 trillion market could be allocated to generative AI computing, networking and storage infrastructure.

AI servers are expected to dominate spending, accounting for roughly 80-85% of the total addressable market, or about $700 billion-including approximately $650 billion in accelerators. Networking is projected to capture 10-15%, or $74 billion, while storage will represent around 5%, or $39 billion.

By 2030, total spending on AI data center systems is forecast to exceed 10% of global IT expenditures and make up more than 80% of all data center infrastructure outlays.

"We expect every major country/region to invest in creating its independent sovereign AI factory, trained in local language, culture, generating high-tech employment, and serving critical healthcare, defense, industrial, financial and cyber needs," Arya said.

According to Bank of America Nvidia is expected to retain more than 80% of the AI accelerator market over time, while AMD is projected to hold around 5%.

"Nvidia remains key beneficiary, but the rising tide is also positive Broadcom, AMD, Marvell, optical, foundry, memory, and semicap peers," he added.