Breaking

Apple Weighs Enterprise AI Server Return With M8 Ultra Chips and Nvidia NVLink Fusion

September 19, 2026 at 10:31 EDT

Infra & Chips
Sales & Business
Foundation Models

Apple is reportedly considering a return to the server market with hardware designed for companies and governments that want to run AI models on their own machines instead of relying on public clouds. The system is described as an AI inference-focused server built around future Apple M8 Ultra processors, with Apple also said to have discussed using Nvidia NVLink Fusion as part of the machine's connectivity stack.

September 2026 · Apple

Apple studies AI servers for customer-controlled deployments

A reported enterprise server plan would pair future M8 Ultra processors with possible Nvidia NVLink Fusion connectivity, aimed at organizations that want AI inference on their own machines rather than public clouds.

Reported chip configurations

2M8 Ultra chips

4M8 Ultra chips

The higher-end option doubles the reported processor count.

2029

Reported target window

2 or 4

M8 Ultra configurations

On-site

Customer-controlled AI inference

What the system would try to solve

Control

Run models on owned machines.

Privacy

Keep sensitive data closer to the organization.

Scale

Use high-speed interconnects for AI infrastructure.

Why it attracts interest

Apple silicon could give privacy-sensitive customers another path for local AI inference.

Why caution remains

The product is unannounced, the Nvidia role is unconfirmed, and the reported timing is years away.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.