Apple

Apple is pressing the acquisition accelerator for its AI server chip strategy. On July 15, according to The Information, Apple is actively seeking chip company acquisition targets and has engaged with investment banks and several semiconductor startups over the past few months to assess potential deals. Previously, Apple’s self-developed next-generation AI server chip Baltra encountered delays, while the current M2 Ultra has hit a performance ceiling in data center AI tasks, forcing Apple to outsource heavy workloads to Google Cloud’s Nvidia infrastructure.

ai, marking the second-largest acquisition in the company’s history, second only to the $3 billion acquisition of Beats Electronics in 2014. Apple’s current AI data center computing power relies on the M2 Ultra chip, but as the scale of AI workloads grows, its internal servers have reached a performance bottleneck. According to The Information, Apple’s next-generation AI server chip Baltra was originally scheduled to debut this year, but has been delayed. Before Baltra is delivered, Apple has no choice but to outsource some heavy AI tasks to external infrastructure.

These workloads are currently running on Nvidia chips through Google Cloud. This situation means that, as AI models continue to iterate rapidly week by week, Apple’s most core AI infrastructure is now doubly dependent on external suppliers. Notably, a recent Morgan Stanley research report revealed that Apple is locking in large amounts of TSMC’s SoIC advanced packaging capacity for Baltra—36,000 units ordered for 2026, and a surge to 60,000 units in 2027. ai this January marked an inflection point in its acquisition strategy. ai focuses on technology that interprets speech via subtle facial movements.

The nearly $2 billion valuation indicates Apple is no longer bound by its previously restrained approach toward strategic acquisitions. The rationale behind this deal is not just about acquiring technology, but also about demonstrating a path forward: when internal R&D falls behind expectations, external acquisitions can quickly fill capability gaps. Now, this logic is being replicated in the AI server chip sector.