Apple May Return to Server Market with Nvidia Technology
2 hours ago
- Apple is reportedly developing an AI server using its M8 series chips and considering Nvidia's NVLink Fusion networking hardware for data exchange between chips.
- The server would target companies wanting to run AI models (especially inference) on their own equipment, potentially reviving Apple's server business discontinued in 2011 with the Xserve.
- A potential release date is 2029, but the project may be canceled; Apple's CEO John Ternus supported the initiative when it began about a year ago.
- Apple already produces server hardware for Private Cloud Compute, but scaling current chip connections poses cost and speed issues, leading to consideration of Nvidia's technology.
- An agreement with Nvidia would give the company a role in an Apple product that could compete with its own AI systems, while NVLink Fusion allows Nvidia to supply networking to alternative processor makers.
- The project could improve Apple-Nvidia relations after past disputes; Apple has already extended Private Cloud Compute to Google Cloud using Nvidia GPUs and explored other AI collaborations.
- Challenges remain, including the need for stronger business customer support, more AI developer software resources, and further investment in Apple's MLX framework.