Apple Inc. is exploring the acquisition of semiconductor companies to strengthen its internal AI server-chip capabilities [1, 2].
This move signals a critical gap in Apple's hardware strategy as it attempts to compete in the generative AI race. While the company emphasizes vertical integration, its current in-house silicon is reportedly unable to meet the performance demands of large-scale AI models [1, 2].
Reports indicate that the M2 Ultra chip struggles with AI workloads, which has led to delays in the development of the next-generation server chip, codenamed "Batra" [1, 2]. To bridge this performance gap, Apple has relied on external infrastructure. The company is currently using Nvidia chips via Google Cloud to power its high-performance tasks [1, 2].
This reliance on competitors comes as Apple pushes forward with software integration. At the Worldwide Developers Conference (WWDC) held in June 2024, the company debuted a version of Siri integrated with Google's Gemini [1, 2].
To resolve these hardware limitations, Apple is pursuing discussions with various semiconductor startups [1, 2]. By acquiring specialized expertise, the company aims to accelerate the timeline for its own AI silicon, and reduce its dependency on third-party cloud providers and hardware manufacturers like Nvidia [1, 2].
“Apple is exploring acquisitions of semiconductor companies to strengthen its AI server‑chip capabilities”
Apple's shift toward semiconductor acquisitions reveals a vulnerability in its 'Apple Silicon' ecosystem. While the company successfully transitioned Macs to its own chips, the massive compute requirements of Large Language Models (LLMs) have outpaced the M2 Ultra's capabilities. By relying on Nvidia and Google Cloud, Apple is temporarily conceding the hardware advantage to its rivals while it attempts to buy the expertise necessary to build a competitive AI server infrastructure.
