Apple's updated desktop portfolio officially landed on retail shelves and kicked off worldwide customer deliveries. The rollout brings the redesigned, compact Mac mini—powered by the new M6 and M5 Pro processors—and the workstation-class Mac Studio, featuring M5 Max and M5 Ultra silicon, directly to physical Apple Stores and authorized retail channels.
The transition marks an architectural inflection point for desktop Apple silicon, integrating dedicated Neural Accelerators straight into each GPU core and introducing multi-system clustering capabilities over Thunderbolt 5 for heavy enterprise and local machine learning pipelines.
Mac mini: M6 Efficiency and M5 Pro Performance Tiers
The compact Mac mini continues to anchor Apple's entry desktop tier while raising baseline computing power. Built on a refined 3-nanometer node, the base M6 configuration pairs a 12-core CPU with a 12-core GPU and an upgraded Dual 16-core Neural Engine, yielding up to 40 percent faster CPU execution and double the graphics bandwidth over the previous generation.
Features a 12-core CPU and 12-core GPU with unified memory starting at 16GB (170GB/s bandwidth), configurable up to 32GB for everyday productivity and creator workflows.
Scales up to an 18-core CPU, a 20-core GPU, and up to 64GB of unified memory operating at 307GB/s, targeted at heavy compile tasks and multi-track audio production.
Both Mac mini tiers come standard with Wi-Fi 7, Bluetooth 6, and 2.5Gb Ethernet out of the box, with a 10Gb Ethernet port available as a build-to-order option.
Mac Studio: M5 Ultra Silicon & Clustered Workflows
At the workstation level, the refreshed Mac Studio targets heavy-duty rendering, continuous 8K video timelines, and local large language model inference. The M5 Max configuration provides an 18-core CPU, an up-to-40-core GPU, and up to 128GB of memory.
The flagship M5 Ultra doubles the compute array, scaling up to a 36-core CPU, an 80-core GPU, and a massive 512GB unified memory ceiling pushing 1.2TB/s of bandwidth. Both workstation configurations introduce 120Gb/s Thunderbolt 5 ports supporting Remote Direct Memory Access (RDMA), enabling direct node-to-node machine clustering for distributed local AI workloads.
Hardware Architecture & Pricing
| Hardware Tier | Mac mini (M6 / M5 Pro) | Mac Studio (M5 Max / M5 Ultra) |
|---|---|---|
| Target Segment | Compact Desktop / Prosumer | Pro Workstation / AI Labs |
| CPU Configurations | 12-Core (M6) / Up to 18-Core (M5 Pro) | 18-Core (M5 Max) / Up to 36-Core (M5 Ultra) |
| GPU Configurations | 12-Core (M6) / 20-Core (M5 Pro) | Up to 40-Core (M5 Max) / Up to 80-Core (M5 Ultra) |
| Max Memory Capacity | Up to 64GB Unified Memory | Up to 512GB Unified Memory |
| Max Memory Bandwidth | Up to 307 GB/s | Up to 1.2 TB/s |
| High-Speed I/O | Thunderbolt 4 / Optional 10GbE | Thunderbolt 5 (120Gb/s) with RDMA Clustering |
| Starting Price | $899 (M6) / $1,699 (M5 Pro) | $2,499 (M5 Max) / $5,499 (M5 Ultra) |
Operating System Integration
Every configuration across the Mac mini and Mac Studio lineups ships preloaded with macOS 27 Golden Gate, incorporating deeper systemwide Siri AI hooks, context-aware Visual Intelligence across open windows, and automated tab triage within Safari.
The headline move here is hardware-level clustering over Thunderbolt 5. Instead of forcing enterprise users and researchers into dedicated server racks, Apple is turning multiple linked Mac Studios into modular, low-latency AI inference arrays right on the desk.
Sources: Official release details via Apple Newsroom with hardware validation via MacRumors.
