Apple is reportedly developing an enterprise server powered by its own chips and has discussed using Nvidia networking technology to connect them, according to The Information.
The company is considering selling the system to AI developers, businesses and government customers, which could mark Apple’s return to the commercial server market after more than a decade.
Up to Four M8 Ultra Chips
Apple is reportedly considering two versions of the server.
The smaller model would feature two upcoming M8 Ultra chips, while a higher-end configuration would combine four M8 Ultra chips.
Apple has discussed linking the processors using Nvidia’s NVLink Fusion, a high-speed interconnect technology that Nvidia also licenses to companies including Qualcomm, Arm, Fujitsu and Marvell.
However, the project is still in development. The server is reportedly not expected to launch before 2029, and Apple could ultimately cancel the project or release it without Nvidia’s networking technology.
Shares of Apple and Nvidia showed little movement following the report.
Apple Has Not Sold a Server Since 2011
Apple already operates servers for its own infrastructure.
Its Private Cloud Compute system initially relied on Apple silicon. In June, the company also expanded parts of its AI infrastructure through Google Cloud using Nvidia GPUs and Intel processors.
However, Apple has not sold a dedicated server product since 2011, when it discontinued the Xserve after nearly nine years on the market.
The reported M8 Ultra system would therefore represent a return to selling server hardware directly to external customers.
Governments Could Become an Important Market
The possibility of government customers comes as Europe places greater emphasis on technological sovereignty.
The European Commission proposed its Cloud and AI Development Act in June. The proposal sets out four sovereignty assurance levels for cloud services used by public-sector organizations, with requirements becoming stricter as the sensitivity of workloads increases.
At the highest level, providers would face extensive requirements for software supply-chain control and transparency while also having to prevent interference from third countries.
Europe has already indicated that relying on non-European technology does not automatically prevent a service from meeting sovereignty requirements.
In April, the European Commission awarded a six-year sovereign cloud framework worth up to €180 million to four European providers. One winning group was led by Proximus and included S3NS, a joint venture between Thales and Google Cloud, along with Clarence and Mistral.
The bidders were assessed against eight sovereignty objectives, including legal exposure, supply-chain transparency, technological openness, security and operational control.
Hardware Falls Outside the Proposed Cloud Rules
The proposed Cloud and AI Development Act focuses on cloud services rather than server hardware. As a result, a direct Apple server sale would not itself be subject to those specific cloud-service requirements.
Governments, however, can establish their own procurement requirements when purchasing infrastructure.
This creates a distinction between sovereign technology and sovereign operation. A government-owned server running within its own facilities could face a different sovereignty assessment from a cloud service operated by an external provider.
For Apple, that could make locally operated enterprise hardware relevant to governments seeking greater control over their AI infrastructure, even if the underlying processors and networking technology remain American.
For now, the project remains unconfirmed, with no commercial launch expected before 2029.
