Apple expands AI server ambitions with new chip and data center plans

Bloomberg's Mark Gurman reports that Apple is accelerating its in-house enterprise AI server and chip initiatives. One of the self-developed chips may be based on the M7Ultra and is set for release in 2029, aiming to compete with NVIDIA. Apple plans to launch two server versions, with the base model featuring two M8Ultra chips and the high-end model with four. The company is also exploring NVIDIA's NVLink Fusion technology to enhance chip collaboration in data centers. Gurman notes Apple is set to introduce a new AI server by early 2027, marking a shift from terminal chip design to data center infrastructure. The Information previously reported Apple's intention to use its own AI chips in enterprise servers. Specific details and release dates may still change.