Apple's R&D investments top 10% of sales as AI race creates 'sense of urgency'

Don't confuse R&D with CapEx

Apple’s R&D spending hit 10.3% of revenue in the March quarter, as the the tech giant ramps up investments in artificial intelligence.

The increase puts Apple closer to the rest of megacap tech on R&D, though the iPhone maker is still taking a very different approach when it comes to capital expenditures, according to CNBC. R&D spending was up from 7.6% in the prior period and 9% in the same quarter a year earlier. While sales jumped 17%, the fastest rate of growth for any quarter since 2021, R&D climbed twice as quickly, increasing almost 34% from a year prior, according to CNBC.

“That’s a sign that Apple is seeing a sense of urgency around new AI products,” said Gene Munster, managing partner at Deepwater Asset Management.”

Horace Dediu, founder of Asymco, agreed that a major question is where all that spending is going. Much of the increase, he said, is likely tied to talent, teams and experiments in training and modeling, rather than the kind of large-scale data center deployment underway at the hyperscalers. But he also said Apple is pushing across hardware, investing in silicon, optics, batteries, materials, sensors and smaller form factors.”

"AI is an obvious candidate, but R&D is not capex," Dediu said.

Excerpt From Apple's R&D investments top 10% of sales as AI race creates 'sense of urgency' by MacKenzie Sigalos, Jennifer Elias, CNBC

Source: CNBC

NB: Apple has been increasing its R&D spending from around 2% to 9% of sales on a 4-trailing quarter basis. It has overtaken SG&A which has remained around 7% since 2011.

This is a big departure from where the company was during the Steve Jobs era and the acceleration seems to be continuing. An increase to 10% seems very likely this year with a run rate of $40 billion/yr.

This level of R&D used to be associated with companies like Microsoft and Google which had dedicated laboratories. The question is where does this spending go to?

AI is an obvious candidate but R&D is not CapEx. Although some capital equipment is part of the spend, much of this is talent related as it’s recurring. My assumption is that there are many projects that are both hardware and software (e.g. Apple Silicon and AI). Fundamentally the power of this budget is affecting the organizational behavior with, as we have seen, the ascent of engineering leadership.

AI is expensive but so is silicon and so are the components going into smaller form factors. They are pushing on many fronts: optics, batteries, materials, sensors and silicon.

AI spending is teams and experiments in training and modeling. It would cost a lot more to deploy these at scale and as much as this is a lot of spending, it’s not nearly what super scaling requires.

The late 90s to early 2000s period was defined by revenue contraction and financial distress. It wasn’t so much an increase in R&D as a decrease in revenues. In those days Apple built products on a shoe string. Literally a handful of people were tasked with scoping out new products. The iPod was Tony Fadell and a few others. Safari effort started literally with one hire. I don’t think that era is comparable to what we are seeing today. There must be thousands of engineers allocated to R&D right now.