Apple is weighing the use of Nvidia technology as part of a possible return to the server market, according to a report from The Information cited by Reuters on Wednesday.
The report suggests the iPhone maker has at least explored what it would take to build and sell server hardware again â a business it exited more than a decade ago â and that Nvidia’s technology could feature in those plans. Neither Apple nor Nvidia has publicly confirmed the deliberations, and reports of internal exploration at Apple frequently describe projects that never reach the market.
A business Apple walked away from
Apple was once a genuine, if modest, participant in the server business. Its Xserve rack-mounted machines and the Mac OS X Server software that ran on them found a following among creative studios, universities and small businesses through the 2000s. Apple discontinued the Xserve line at the start of the 2010s, steering customers toward Mac Pro and Mac mini configurations instead and effectively ceding the data center to Dell, HP and a long list of white-box builders.
In the years since, Apple has focused its hardware ambitions on consumer devices and on designing its own silicon. The transition of the Mac to Apple’s in-house chips, and the extension of that silicon work into custom accelerators and networking components, has given the company a far deeper hardware engineering bench than it had when it abandoned servers.
Why now
The context for any return is obvious: the boom in demand for computing capacity to train and run artificial intelligence models. Nvidia has been the dominant supplier of that capacity, and its accelerators and associated systems technology sit at the center of most large-scale AI infrastructure being built today. Any company contemplating server hardware with AI workloads in mind would have reason to evaluate Nvidia’s offerings.
Apple has also been expanding its own infrastructure footprint. The company has built out capacity to support cloud-based AI processing for its devices, including servers that run Apple’s own silicon in a privacy-focused configuration. That work has already forced Apple to think seriously about data center-class hardware â a stepping stone, potentially, to selling such hardware or offering services built on it.
Questions the report leaves open
A great deal remains unclear. It is not evident from the reporting whether Apple is contemplating selling servers to enterprise customers, building systems purely for internal use, or something in between, such as offering compute capacity as a service. Nor is it clear how Nvidia’s technology would be combined with Apple’s own chip designs, which the company has invested heavily in and generally prefers over third-party silicon in its products.
The competitive picture is daunting. The market for AI servers is crowded with entrenched vendors and with cloud providers that design their own systems at enormous scale. Margins in commodity server hardware are thin by Apple’s standards, and enterprise sales require a support and channel apparatus that Apple has spent years de-emphasizing.
Still, the report is a reminder of how thoroughly the AI buildout has redrawn the boundaries of the technology industry. Companies that once had no interest in data centers are reconsidering, and a partnership between Apple and Nvidia â two firms that have not always been close collaborators â would be a notable signal in itself.
For now, it remains a report of internal deliberation, not an announcement. Investors and enterprise buyers will have to wait for Apple itself to say more. Read More
Leave a Reply