Apple Doubles Down on AI While Keeping Capital Spend Lean

Apple closed its fiscal year with the highest ever revenue and earnings per share, driven by a 17.5% jump in iPhone sales to $57.6 billion and a 24% rise in Mac revenue to $10.8 billion. The company cautioned, however, that rising memory prices will compress margins in the holiday quarter.

During the earnings call, AI and "Apple Intelligence" were referenced 45 times — a record — signaling that executives view the technology as a primary catalyst for future upgrades, subscription fees, and third‑party app partnerships.

Unlike its peers, Apple's capital expenditure remained modest at $2.2 billion last quarter, far below Amazon's $77.3 billion and Alphabet's $51.4 billion. The firm blends on‑device inference, its own server fleet, and external models such as Google's Gemini, keeping infrastructure spend a fraction of the industry average.

Why it matters for GPU / AI infrastructure

Apple's lean capex approach means it does not directly drive massive GPU cluster purchases, but the growing mix of on‑device and cloud‑based inference — especially as Siri AI expands to more languages — will still create demand for high‑performance GPUs at the edge and in partner clouds.

The revamped Siri launched in beta for U.S. English in September and is rolling out to French, German, and other locales. As usage scales, the associated cloud costs will appear in services cost of sales, offering a measurable indicator of AI‑driven infrastructure consumption.