Apple is actively pursuing acquisitions of semiconductor startups to shore up its in-house server chip development for AI workloads. The company has engaged investment bankers and reached out to various chip startups to gauge interest in potential deals, though no specific targets have been named publicly.

Apple’s M2 Ultra chips, which the company has been deploying in its AI server infrastructure, have shown performance shortfalls when handling heavy AI workloads. Apple’s goal with these potential acquisitions is to reduce its dependence on Nvidia for high-performance AI tasks.

The current roadmap involves using M2 Ultra and M4 chips for near-term AI server applications, with future plans pointing toward advanced custom chip designs built specifically for AI workloads from the ground up.

This isn’t Apple’s first time buying its way into chip dominance. The company acquired PA Semi back in 2008 for roughly $278 million, a deal that laid the groundwork for the A-series chips that eventually powered every iPhone. Apple has since applied a similar pattern with laptop and desktop chips via the M-series transition away from Intel.

For semiconductor stocks, companies that Apple might acquire could see premium valuations, while Nvidia’s dominance narrative faces another credible challenger. Every hyperscaler that builds custom AI chips is one more customer that isn’t buying Nvidia’s most expensive products.