Apple Considers Nvidia Tech for Return to Server Market: The Information

Apple is reportedly preparing to make a potential return to the enterprise server market, more than a decade after discontinuing its Xserve server line.

According to a report by The Information, Apple is developing an artificial intelligence server designed for AI developers, businesses and government customers. The company is considering using its own high-performance M8 Ultra chips in the system and has also discussed incorporating Nvidia's NVLink Fusion technology.

The reported move would represent a significant expansion of Apple's custom silicon strategy beyond Macs and other consumer devices. Reuters, citing The Information's report, said the proposed system is aimed primarily at AI inference workloads rather than AI model training.

Apple Exploring a Return to the Enterprise Server Market

Apple exited the server hardware market in 2011, when it discontinued its Xserve product line.

The company subsequently concentrated its hardware strategy on consumer and professional products such as Macs, iPhones, iPads and other devices, while continuing to develop increasingly powerful proprietary processors.

The reported server project would mark a potential return to enterprise hardware, this time with a strong focus on artificial intelligence.

According to The Information, Apple has been working on the project for about a year. The initiative is reportedly backed by Apple CEO John Ternus, although the project could still be modified or cancelled before reaching the market.

M8 Ultra Chips Could Power Apple's AI Servers

The proposed server could reportedly be offered in configurations featuring two or four M8 Ultra chips.

The M8 Ultra is described in the report as Apple's planned highest-performance processor. Combining multiple high-end Apple-designed chips could allow the company to create a system capable of handling demanding AI workloads.

Apple has increasingly used its own silicon to control the performance, power efficiency and integration of its hardware products.

The reported server initiative would extend that approach into enterprise computing, where demand for high-performance AI infrastructure has grown rapidly.

Why Apple Is Considering Nvidia's NVLink Fusion

One of the most notable aspects of the reported project is Apple's consideration of Nvidia's NVLink Fusion technology.

NVLink is Nvidia's high-speed interconnect technology designed to allow processors and accelerators to communicate efficiently. NVLink Fusion is intended to enable third-party custom processors to connect with Nvidia's broader data-centre ecosystem.

For Apple, incorporating Nvidia networking technology could potentially make its custom silicon more compatible with infrastructure already built around Nvidia's AI computing technologies.

Reuters reported that the use of NVLink Fusion would allow Apple's custom AI chips to better interface with Nvidia's data-centre infrastructure.

Apple's Own Chips and Nvidia Networking

The reported plan would create an interesting combination of technologies.

ComponentReported TechnologyProcessorApple's M8 UltraPossible configurationTwo or four M8 Ultra chipsNetworking/interconnectNvidia NVLink FusionPrimary workloadAI inferenceTarget customersAI developers, businesses and governmentsPotential launch2029Current statusDevelopment; subject to change or cancellation

The approach would allow Apple to retain control over its core processing hardware while potentially using Nvidia technology for high-speed connectivity.

AI Inference Is a Key Focus

The proposed Apple server is reportedly designed primarily for AI inference.

AI inference refers to the process of using a trained AI model to generate outputs, such as responses from a chatbot, predictions, classifications or other results.

This differs from AI training, which involves using large datasets to develop and adjust an AI model.

As companies deploy AI applications at scale, demand for inference infrastructure has become an increasingly important part of the data-centre market.

Apple's reported focus on inference suggests that the company could be targeting businesses and organisations looking to operate AI workloads rather than directly competing across every segment of the AI training market.

Apple Could Target Businesses and Governments

Unlike Apple's traditional consumer hardware, an enterprise AI server would target a different customer base.

The Information report says the proposed system could be aimed at:

• AI developers

• Businesses

• Government organisations

• Enterprise technology users

• Organisations requiring local AI inference infrastructure

Enterprise customers typically consider factors such as performance, security, deployment flexibility, energy consumption and compatibility with existing infrastructure when purchasing servers.

Apple's ecosystem and custom-chip expertise could therefore be applied to a different segment if the project eventually reaches commercial production.

What Is NVLink Fusion?

NVLink Fusion is part of Nvidia's technology for connecting custom processors with Nvidia's high-performance computing ecosystem.

The technology could become relevant for companies developing their own application-specific processors while still requiring high-speed connectivity to Nvidia-based infrastructure.

Apple's reported interest is notable because the company has historically emphasised its own hardware and software technologies.

Using Nvidia's interconnect technology would not necessarily mean Apple is abandoning its custom silicon strategy. Instead, it could represent a combination of Apple's processors with Nvidia's networking and data-centre technologies.

A Potential Shift in Apple's AI Hardware Strategy

Apple has invested heavily in developing its own chips for Macs and other products.

The company has used its custom silicon strategy to integrate CPUs, GPUs, neural processing capabilities and other components into unified systems.

A dedicated AI server could represent another stage in that strategy.

Rather than relying entirely on third-party processors for enterprise AI systems, Apple could potentially use its own silicon as the foundation of a server platform.

At the same time, the reported consideration of Nvidia's NVLink Fusion shows that Apple may be willing to use specialised external technologies where they could improve interoperability or performance.

Apple's Previous Server Business

Apple's potential return to servers would be notable because the company previously had a dedicated server product line.

Apple discontinued Xserve in 2011, ending its direct server hardware offering.

Since then, enterprise customers have generally relied on Macs, software services and third-party server infrastructure rather than a current Apple-branded enterprise server platform.

The reported AI server project would therefore represent a new approach rather than simply a revival of the old Xserve strategy.

Why the Server Market Is Changing

The server market has been transformed by the rapid expansion of artificial intelligence.

Traditional enterprise servers have increasingly been supplemented by specialised AI infrastructure designed for demanding workloads.

Companies developing AI systems need computing resources for both training and inference. These systems also require high-speed networking because AI workloads can involve communication between multiple processors and accelerators.

This has created demand for specialised chips, interconnect technologies, memory and data-centre infrastructure.

Apple's reported project appears to be focused on this newer AI-driven portion of enterprise computing.

Potential Relationship Between Apple and Nvidia

Apple's reported consideration of Nvidia technology is also notable because the relationship between the two companies has had periods of tension.

The two companies have worked within the same technology ecosystem for years, but Apple's move toward proprietary silicon has reduced its reliance on Nvidia graphics processors in many of its products.

The reported consideration of NVLink Fusion could nevertheless indicate a more practical relationship in the data-centre market.

Rather than using Nvidia's processors as the main computing engine, Apple could potentially use Nvidia technology for connectivity while maintaining its own processors at the centre of the system.

Reuters noted that the move could also reflect improving relations between Apple and Nvidia after years of friction related to hardware issues.

Potential 2029 Launch

The proposed Apple AI server is reportedly tentatively scheduled for release in 2029.

However, that timeline should not be treated as a confirmed launch date.

The project is still under development, and Reuters reported that it could be altered or cancelled. Apple and Nvidia had not commented on the report at the time of publication, and Reuters said it had not independently verified the information.

Therefore, the reported 2029 timeframe represents a potential target rather than a confirmed Apple product announcement.

What Apple's Server Move Could Mean for Enterprise AI

If Apple eventually launches an AI-focused enterprise server, it could broaden the competitive landscape in AI infrastructure.

The company would potentially bring its expertise in:

• Custom processor design

• Hardware-software integration

• Energy-efficient computing

• AI processing

• System-level optimisation

into an enterprise environment.

The use of Nvidia's networking technology could also allow Apple to address compatibility requirements in data-centre environments where Nvidia technology is already widely deployed.

However, the commercial significance of such a product would depend on pricing, performance, software compatibility, availability and Apple's enterprise support strategy.

Apple's Broader AI Push

The reported server project comes as Apple continues to increase its focus on artificial intelligence.

The company's proprietary processors include dedicated capabilities for AI-related workloads, while Apple has also been developing AI features across its consumer product ecosystem.

An enterprise server would potentially give Apple another route to deploy its silicon for AI applications beyond smartphones, tablets and personal computers.

Such a move could also allow Apple to participate more directly in the rapidly expanding market for AI computing infrastructure.

Key Takeaways

• Apple is reportedly developing an enterprise AI server.

• The system could use two or four M8 Ultra chips.

• Apple is reportedly considering Nvidia's NVLink Fusion technology.

• The server is primarily intended for AI inference workloads.

• Potential customers include AI developers, businesses and governments.

• Apple exited the traditional server market after discontinuing Xserve in 2011.

• The reported project has been in development for around a year.

• A 2029 launch has reportedly been considered, but the timeline is not confirmed.

• The project could still be modified or cancelled.

• Apple and Nvidia have not publicly confirmed the reported server plans.

Bottom Line

Apple is reportedly exploring a return to enterprise servers, but with a distinctly AI-focused strategy.

The proposed system could combine Apple's M8 Ultra processors with Nvidia's NVLink Fusion technology, allowing Apple to use its own custom silicon while potentially connecting to Nvidia's data-centre ecosystem.

The reported project reflects the changing nature of enterprise computing as AI inference becomes an increasingly important workload. However, Apple has not confirmed the project, and the reported 2029 launch timeline remains tentative.

Disclaimer

This article is based on reports from The Information and Reuters and describes a product reportedly under development. Apple has not publicly confirmed the reported server project or launch timeline. Product specifications, launch plans and technology partnerships may change before any commercial announcement.