Apple is reportedly exploring acquisitions of AI-chip companies to strengthen the server infrastructure behind Apple Intelligence, a notable shift for a company better known for buying small teams and technologies than pursuing multibillion-dollar deals.
The Information reports that Apple has spoken with bankers and semiconductor startups as it looks for expertise that could accelerate custom chips for AI workloads in its data centers. The effort reportedly follows limits with Apple’s existing M2 Ultra-based server systems when handling larger models.

Futuristic data center with quantum computing, microchips, GPUs, global networks, and business collaboration.A server-side gap, not a new Mac chip​

Apple Silicon has been highly successful in phones, tablets and PCs, where power efficiency and tight hardware-software integration matter most. Training and serving large AI models in data centers is a different job: it demands massive memory capacity, fast interconnects and software ecosystems built around clusters of accelerators.
According to The Information, Apple has relied on Nvidia hardware in Google Cloud for some of the more demanding work associated with its next-generation Siri plans, including use of Google’s Gemini models. That does not mean Apple has abandoned in-house silicon; it means the company’s present server hardware reportedly cannot cover every workload it wants to run itself.
A future Apple AI server processor, reportedly code-named Baltra, has also slipped beyond its planned 2026 arrival. Apple has not publicly confirmed the chip, its schedule, or any acquisition targets.
For Apple, acquiring a specialist could be faster than building every component internally. It also would echo the company’s 2008 purchase of PA Semi, which helped seed the internal silicon organization that later produced the A-series and M-series processors.

Bigger deals appear more plausible​

Apple’s reported purchase of Israeli AI startup Q.ai for close to $2 billion in January was its second-largest acquisition after the 2014 Beats deal. While Q.ai focused on interpreting speech through facial micromovements rather than silicon, the price suggested Apple may be more willing to spend heavily where it sees strategic value.
There is also a concrete sign that Apple is extending its chip supply relationships. Broadcom disclosed in a July 6 regulatory filing that it and Apple had expanded their long-standing technology collaboration through 2031, covering custom ASIC products for multiple generations of Apple devices. Apple separately said the expanded arrangement would support Broadcom manufacturing investment in Colorado.
That agreement does not establish that Broadcom is building Apple’s AI server processor. But it reinforces that Apple’s AI infrastructure plans will not be limited to chips designed wholly inside Cupertino.

Why Windows users and admins should care​

There is no Windows update, compatibility change, or enterprise action tied to this report. The practical relevance is market-wide: Apple is joining Microsoft, Google, Amazon and Meta in treating AI compute as a core infrastructure capability rather than merely a cloud service purchased from Nvidia.
For Windows administrators, Nvidia’s CUDA ecosystem and the broader Windows AI PC stack remain unchanged. But if Apple begins buying AI-silicon talent or companies, competition for accelerator supply, engineering staff and data-center capacity could intensify further.
For now, the story is a reported strategy exploration, not an announced acquisition or product roadmap.

References​

  1. Primary source: AI: Reset to Zero
    Published: 2026-07-19T05:02:31+00:00
  2. Official source: apple.com
  3. Related coverage: axios.com
  4. Related coverage: techradar.com