Breaking

Apple Reportedly Plans M7 Ultra-Based AI Server for 2029 as It Pushes Custom Silicon Deeper Into the Cloud

September 16, 2026 at 10:09 EDT

Infra & Chips
Foundation Models

Apple is reportedly developing an enterprise AI server built around its own high-end silicon, with a follow-on system tied to the M7 Ultra chip envisioned for 2029. The effort would extend Apple Silicon beyond Macs and devices into heavier cloud AI infrastructure, as the company tries to support Apple Intelligence while reducing reliance on outside accelerators.

2029 · Apple

Apple's AI server bet moves toward custom cloud silicon

Reports point to an M7 Ultra-based enterprise AI server around 2029, extending Apple Silicon from devices and Macs into heavier cloud inference infrastructure.

2029

Reported server target

1.5TB

M7 Ultra memory goal

On-device model parameters

1.15x

Research cost-efficiency result

The memory target is the headline spec

1.5TB15 blocks at 0.1TB each

A large unified-memory pool would be aimed at memory-heavy AI inference, but reports say the configuration may depend on DRAM supply conditions.

M5 Ultra server

Initial Apple AI server scaling

→

M7 Ultra chip

High-memory AI silicon target

2029 server

Reported enterprise AI system

Why it matters

Apple could align devices, models, cloud serving and chips under one AI stack.

Main caveats

The 2029 target is distant, specifications are unconfirmed, and software ecosystems remain a major factor.

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.