Apple plans to return to the server market with Nvidia technologies

Apple is considering a return to the server market by integrating Nvidia hardware solutions. The company's strategy involves creating specialized computing nodes to support its own cloud services and machine learning tasks.

According to internal reports, the company plans to use the latest generation of graphics processors to scale the infrastructure needed for training and inference of its own generative models. The transition to a high-density computing architecture will reduce latency when processing user requests within the Private Cloud Compute ecosystem.

Previously, the company discontinued the production of Xserve servers in 2011, focusing on consumer devices. The new initiative aims to reduce dependence on third-party cloud providers and optimize the costs of training neural networks, which require significant power for parallel computations.

The integration of Nvidia chips into Apple's server architecture will be a key step in the development of hardware for AI. The deployment of the first clusters based on the new servers is expected to begin within the next 18 months, allowing the company to control the entire data processing cycle from the user device to the server center.

This move marks a change in the approach to data management, where the hardware part of the servers will be optimized for the specific requirements of Apple's neural network algorithms. This will enable the company to improve the performance of its AI functions while ensuring a higher level of privacy due to the localization of computing power.