The news
Apple announced a new Mac mini configurable with either the new M6 chip or the M5 Pro chip released earlier this year. The base model starts at $899. The M6 carries a 12-core CPU and 12-core GPU, compared with the 10-core CPU and 10-core GPU in the prior M5. The M6 also introduces the first dual Neural Engine design, pairing two 16-core engines.
Context
The previous Mac mini generation shipped with the M4 chip. That model set the baseline for the smallest and cheapest desktop Mac in Apple’s lineup. The update replaces the M4 entry point while adding the M5 Pro as a higher-tier option already used in current MacBook Pro models. Buyers now select between the new M6 for entry-level performance or the M5 Pro when they need more GPU cores and higher memory ceilings without moving to a Mac Studio or Mac Pro.
The change keeps the Mac mini’s compact form factor and price positioning intact. No chassis redesign or port layout changes appear in the announcement. The focus stays on the silicon upgrade and the performance numbers Apple attached to the M6 configuration.
Details
The M6 chip pairs two 16-core Neural Engines for the first time. Earlier chips used a single Neural Engine. Memory bandwidth increases to 170GB/s from 153GB/s on the M5. Apple states the M6 delivers up to 40 percent faster CPU performance, up to 2 times faster graphics performance, and up to 2 times faster storage speeds than a prior Mac mini with a 10-core M4 chip, 32GB of unified memory, and 2TB of storage. AI workloads see the largest claimed gain, reaching up to 4 times the performance of that reference system.
The M5 Pro option carries over the specifications already shipping in recent MacBook Pro models. Pricing for the new Mac mini begins at $899 for the M6 configuration. The announcement supplies no additional details on thermal design, port changes, or external appearance. All performance figures come from Apple’s internal testing against the specified M4 reference configuration.
Why it matters
Engineers and developers who rely on a compact desktop now receive a clearer entry point for local AI work. The dual Neural Engine and higher memory bandwidth target inference and training tasks that fit inside the unified memory pool. The 4 times AI performance claim, measured against an M4 baseline, points to measurable reductions in time for models that previously ran near the limits of the single Neural Engine.
The storage-speed doubling and graphics improvement matter for workloads that move large datasets or depend on GPU-accelerated tools. Because the Mac mini remains the lowest-cost desktop option, these gains reach users who previously chose between the M4’s constraints or stepping up to a laptop. The continued availability of the M5 Pro gives buyers who need more GPU cores an immediate path without leaving the mini chassis.
The dual Neural Engine architecture appears for the first time in the M6. Whether it scales to future chips or stays limited to this generation will determine how long the advantage lasts. The immediate hardware change, however, is concrete on the specifications Apple released and directly addresses the performance ceiling developers hit with the M4 Mac mini.
---
Sources:
{"word_count": 682, "sources_used": 2}
No comments yet