Discovered AI Shares with Potential for Undervalued Profits:
In the rapidly evolving world of Artificial Intelligence (AI), a handful of less-heralded tech stocks are providing the essential infrastructure that powers the AI gold rush. Among these are Marvell Technology, Credo Technology Group, Camtek, FormFactor, and Vertiv Holdings.
Marvell Technology, a fabless chip designer specialising in wired networking, boasts a broad portfolio including switching, processing, and optical chips. Serving data centers and cloud markets, Marvell is well-positioned to benefit from generative AI investments, particularly due to its optical chips and custom accelerators. Despite recent challenges and a 33% decrease in its stock price in 2025, the company remains 24% undervalued relative to fair value estimates, indicating potential upside.
Credo Technology Group focuses on high-speed connectivity solutions crucial for AI infrastructure. With a significant investment in R&D ($146 million in fiscal 2025, approximately 33% of revenue), Credo is capitalising on the growing demand for AI-related connectivity. This strategic focus has driven a remarkable 221% increase in its stock price from April to July 2025, reflecting its transition into a critical enabler of the AI revolution.
Camtek, an Israeli company that provides inspection and metrology equipment for semiconductor manufacturing, plays a vital role in quality control as chips become more advanced and microscopic. Camtek's solutions for testing High-Bandwidth Memory (HBM) and other AI-specific processors position it well for the coming wave of more powerful AI hardware. In Q1 2025, Camtek reported a 22% year-over-year increase in revenue to $118.6 million, with net income surging 38% to $34.3 million. The company's gross margin hit an exceptional 51%, making it a formidable player in the hardware manufacturing sector.
FormFactor, a company that provides advanced wafer probe cards for verifying functionality of microscopic chip connections, solves the problem of testing chips with thousands of connections operating at high speeds. This challenge is increasingly relevant as AI processors pack more transistors and memory bandwidth. The company recently closed its acquisition of FICT Limited, securing a critical supplier of advanced probe card components. FormFactor's solutions position it well for the future, as the need for advanced inspection grows exponentially as AI chips become more complex.
Vertiv Holdings, meanwhile, provides essential infrastructure for AI data centres. Its liquid cooling systems and precision air conditioning units maintain optimal temperatures, while its uninterruptible power supplies ensure AI training runs without loss of progress due to outages. Vertiv's collaboration with Nvidia and iGenius aims to deploy one of the world's largest AI supercomputers, Colosseum, in Italy. The company reported Q1 2025 results exceeding expectations across various metrics, including a 24% year-over-year increase in net sales and a 25% rise in the company's backlog. Vertiv raised its full-year 2025 sales guidance to $9.3 billion to $9.6 billion, projecting organic growth of 16.5% to 19.5%.
While these companies operate behind the scenes, they form the foundational technology stack enabling AI’s rapid growth. The market has yet to fully recognise the critical roles of these companies in the AI ecosystem, making them compelling, less obvious investments for those seeking to capitalise on the AI boom.
- Marvell Technology, despite facing recent challenges and a decrease in its stock price, remains undervalued due to its essential role in powering the AI gold rush through optical chips and custom accelerators, particularly in generative AI investments.
- Credo Technology Group, by focusing on high-speed connectivity solutions and investing heavily in R&D, is capitalizing on the growing demand for AI-related infrastructure, boosting its stock price and positioning it as a key enabler of the AI revolution.
- FormFactor, through its advanced wafer probe cards for verifying functionality of microscopic chip connections, solves the problem of testing AI processors as they pack more transistors and memory bandwidth, making it a formidable player in the AI hardware sector.
