Apple weighs its first server since the Xserve, and as always, Nvidia may hold the key (www.techradar.com)

Apple is reportedly exploring the possibility of launching its first enterprise AI server since the Xserve, aiming for a potential release in 2029. The new server would utilize two or four of Apple's upcoming M8 Ultra chips and may incorporate Nvidia's NVLink Fusion technology to enhance chip interconnectivity, addressing bandwidth limitations seen with current Thunderbolt-based solutions. This development aligns with the rising demand for high-performance hardware among AI firms—many of which have turned to Apple's Mac minis and Mac Studios for AI training—despite Apple's previous withdrawal from the enterprise server market. The significance of this initiative lies in Apple's unexpected pivot back to a sector it exited over a decade ago due to low sales. With AI companies purchasing Apple’s products in large quantities, the company faces pressure to establish a dedicated AI hardware strategy. The current M5 Ultra's bandwidth capabilities fall short when compared to Nvidia's offerings, which could hinder performance for data-intensive AI tasks. If Apple's plans materialize, their entry into this competitive market could reshape the enterprise landscape, particularly for businesses preferring on-premise AI solutions over public cloud services.