AI Infrastructure Stocks Outpacing Nvidia in 2026

The artificial intelligence sector is experiencing a divergence in 2026, as shares of major hyperscalers and industry leader Nvidia have faced downward pressure while other infrastructure-focused companies report substantial growth [3]. Despite investor skepticism regarding the sustainability of data center spending, more than $700 billion is expected to be invested in AI infrastructure this year [1].

Key takeaways

• Several AI-related stocks, including Bloom Energy, Sandisk, Lumentum, and Micron Technology, have more than doubled in value during the first part of 2026 [2].
• Nvidia shares are down nearly 5% year-to-date, while other semiconductor firms like ASML and Taiwan Semiconductor Manufacturing have posted double-digit gains [3].
• Data center infrastructure providers, such as Vertiv, have seen shares rise by 61.8% so far this year [3].
• Companies are reporting strong returns on AI investments, with some noting that chips are achieving 100% utilization and maintaining high resale value [1].

Infrastructure and Memory Demand Drives Growth

The market shift is largely attributed to the surging demand for components and power solutions necessary to support high-performance data centers. Micron Technology has benefited from the "memory supercycle," as its high-bandwidth memory (HBM) chips are essential for AI performance [1]. Micron reported a 70% year-over-year revenue increase in its 2026 fiscal second quarter, with earnings per share rising to $12.07 [2]. Similarly, Sandisk has seen its revenue grow 251% year-over-year in its fiscal third quarter, driven by a 233% increase in data center revenue [2].

Beyond memory, power and connectivity providers are seeing rapid expansion. Bloom Energy, which provides fuel cell technology for data centers, reported a 130% year-over-year revenue increase in the first quarter of 2026 and achieved profitability with $70 million in net income [2]. Lumentum, a provider of optical products for high-speed data centers, reported a 90% revenue increase for its fiscal third quarter [2]. Meanwhile, companies like AMD are positioning themselves to capture market share in inference and agentic AI, leveraging chiplet designs and high-performance CPUs to meet the evolving needs of AI workloads [1].

Why it matters

The mixed performance of AI stocks suggests that while the market is re-evaluating the "hyperscaler" model, the underlying infrastructure supercycle remains robust [1]. Industry leaders argue that the AI chips currently in use have longer useful lives than initially assumed, and the continued high utilization rates suggest that infrastructure spending is not yet peaking [1]. As companies lock in multi-year contracts for memory and expand their capacity for inference-based AI, the focus has shifted toward the hardware and energy providers that facilitate the broader AI ecosystem [1]. While some analysts remain cautious about the returns on these massive capital expenditures, the companies involved maintain that they would rather risk overspending than miss out on a once-in-a-generation opportunity [1].