Apple Reportedly Explores AI Server Market Return with Nvidia Technology
According to a report by The Information, Apple is considering building enterprise AI inference servers powered by future M8 series chips and Nvidia networking hardware. Targeted for potential release around 2029, these servers would be sold to external customers wanting to run AI models on their own equipment. This move would mark Apple's first return to the hardware server market since discontinuing its Xserve lineup in 2011. It also highlights a potential strategic collaboration between Apple and Nvidia to capture the growing demand for private, on-premises AI inference. Apple is specifically evaluating Nvidia's NVLink Fusion platform to interconnect multiple M8 processors for high-bandwidth data exchange, targeting model inference rather than training. However, the hardware design and partnership are not finalized, and the projected 2029 timeline leaves room for cancellation.
## BACKGROUND
Apple previously competed in the enterprise server space with its Xserve rack-mounted hardware before shutting down the division in 2011 to focus on consumer products. AI inference refers to using an already-trained artificial intelligence model to generate responses or predictions, which requires high-bandwidth interconnect technologies like NVLink Fusion when linking multiple chips together.