Apple has introduced two powerful new members of its Apple silicon family — the M6 and M5 Ultra — promising a major jump in performance, graphics and on-device AI computing.
The announcement, made Tuesday, August 25, puts Apple's desktop Macs at the center of the company's push toward increasingly powerful local AI.
The M6 makes its debut inside the new Mac mini, while the M5 Ultra arrives in the latest Mac Studio. Apple says the chips are designed to handle everything from everyday computing and software development to advanced creative workloads and demanding AI models.
But the biggest story may not simply be raw speed.
Apple is increasingly designing its chips around a future in which more AI processing happens directly on the computer, rather than being sent to remote cloud servers.
⚡ M6: Apple's First 2-Nanometer Mac Chip
The M6 is particularly significant because it is Apple's first chip built using a 2-nanometer process.
The chip features a 12-core CPU, a 12-core GPU and a dual 16-core Neural Engine. Apple says the M6 delivers up to 1.2 times the multithreaded performance of the M5, while its GPU includes dedicated neural accelerators designed to improve AI workloads.
The M6 also supports up to 32GB of unified memory and memory bandwidth of up to 170GB/s.
For everyday users, that could translate into faster application launches, quicker file indexing, faster code compilation and smoother multitasking.
But Apple is putting particular emphasis on AI.
The company says the M6's GPU AI peak compute performance is nearly 30% higher than the M5, while its neural architecture is designed to accelerate on-device AI workloads.
That means developers can increasingly experiment with large language models and AI-powered applications without relying entirely on cloud computing.
🤖 AI Moves Closer to the Mac
The biggest change happening across Apple's silicon strategy is the growing importance of local AI.
Instead of sending every AI request to a remote data center, powerful processors can perform at least some tasks directly on a user's computer.
That has several potential advantages.
Local processing can reduce the need for an internet connection, decrease latency and potentially provide stronger privacy because certain data doesn't need to leave the device.
Apple says developers can use its frameworks, including Core ML and Metal, to take advantage of the new hardware. Developers can also use Apple's foundation models, App Intents and their own AI models to build and run AI workloads on the Mac.
For developers, researchers and AI enthusiasts, that could be one of the most important reasons to pay attention to the M6 and M5 Ultra.
🚀 M5 Ultra: Apple's Most Powerful Chip Yet
If the M6 is aimed at mainstream Mac users and developers, the M5 Ultra is designed for a very different class of workload.
Apple calls it its most powerful chip yet.
The M5 Ultra uses a new quad-die architecture, combining four silicon dies into a single system. It can be configured with up to a 36-core CPU and 80-core GPU, along with up to 512GB of unified memory.
Its unified memory bandwidth reaches 1.2TB/s, giving the processor enormous data throughput for demanding workloads. Apple says that's 50% more memory bandwidth than the M3 Ultra.
This is not a chip designed simply to make web browsing faster.
It is aimed at professional workloads such as:
- Advanced AI development
- Large language models
- 3D rendering
- Video production
- Visual effects
- Scientific computing
- Software development
- Large-scale simulations
🧠Running Huge AI Models Locally
One of the most interesting possibilities with the M5 Ultra is the ability to run increasingly large AI models locally.
Apple says developers can use the enormous unified memory pool and high-speed architecture to work with large language models directly on Mac hardware.
Apple has also discussed connecting multiple Mac Studio systems through Thunderbolt 5 to tackle extremely large AI models. Some configurations could potentially be used to run models containing more than a trillion parameters locally.
That doesn't mean ordinary consumers will suddenly be running trillion-parameter AI models on their desks.
Instead, it shows where Apple believes high-end personal computing is heading: more powerful AI processing at the edge, closer to the user.
🖥️ New Mac mini Gets a Major Performance Upgrade
The M6 isn't arriving alone.
Apple has introduced a new Mac mini powered by M6 and M5 Pro processors. The design remains familiar, but the hardware inside has received significant upgrades.
Apple says the M6 Mac mini can deliver:
- Up to 4× faster AI performance
- Up to 2× faster graphics and storage
- Up to 40% faster CPU performance
The new Mac mini also gains Wi-Fi 7, Bluetooth 6 and faster Ethernet connectivity.
The starting price for the M6 Mac mini is $899, a substantial increase over the previous generation's entry price.
That's an important part of the story.
Apple is offering significantly more performance, but buyers are also being asked to pay more for it.
💻 Mac Studio Targets Professionals
The new Mac Studio is where Apple's M5 Ultra becomes particularly interesting.
The machine continues Apple's compact desktop philosophy while packing extremely powerful hardware inside.
The M5 Ultra configuration starts at $5,499, while the M5 Max Mac Studio starts at $2,499.
That's clearly not a mainstream desktop.
Instead, Apple is targeting professionals who can justify the cost through demanding workloads.
For a filmmaker rendering complex scenes, a developer working with large AI models or a researcher processing enormous datasets, reducing hours of computing time can potentially make a high-end workstation worthwhile.
For someone browsing the web, writing documents and watching videos, however, the M5 Ultra would be serious overkill.
🎮 Graphics Get an Upgrade Too
AI isn't the only focus.
Apple has also upgraded the graphics architecture in the M6.
The chip includes an updated shader architecture, dynamic caching and hardware-accelerated ray tracing.
These technologies can improve rendering performance and gaming graphics, while Apple's neural accelerators can also help with AI-related graphics workloads.
For Mac gaming, the question will be how much of this hardware advantage translates into actual games.
Raw GPU performance is only part of the equation. Game availability, developer optimization and Apple's broader gaming ecosystem will also determine how much users benefit.
📅 When Can You Get the New Macs?
Apple has opened preorders for the new Mac mini and Mac Studio.
The new machines are scheduled to begin shipping on September 22, 2026.
That means buyers won't have to wait long to see how Apple's new silicon performs outside of controlled demonstrations and company benchmarks.
Independent testing will ultimately be important.
Apple's performance figures are useful for understanding its intended improvements, but real-world reviews will tell us how the M6 and M5 Ultra perform across different applications and workloads.
Why This Announcement Matters
Apple's latest chips arrive at an interesting moment for the technology industry.
The AI boom has created enormous demand for computing power and memory. Most of today's biggest AI systems rely heavily on massive cloud data centers packed with specialized hardware.
Apple is taking a complementary approach.
Instead of making every AI experience dependent on the cloud, the company is trying to make personal computers powerful enough to handle increasingly sophisticated workloads themselves.
That could change what people expect from a desktop computer.
A Mac may no longer simply be a machine for running applications. It could increasingly become a local AI workstation capable of running models, processing data and performing sophisticated tasks without sending everything to a remote server.
🔥 The Bottom Line
Apple's M6 and M5 Ultra represent two different directions for Apple silicon.
The M6 is about bringing more performance and AI capability to a relatively compact, mainstream desktop.
The M5 Ultra is about pushing personal computing into workstation and advanced AI territory.
The M5 Ultra is unquestionably the more powerful chip, but that doesn't automatically make it the better choice for everyone.
For most consumers, the M6 Mac mini will be more than enough.
For professional creators, developers and AI researchers working with enormous workloads, the M5 Ultra could be a much more compelling proposition.
The bigger story is what these chips represent.
Apple is betting that the next generation of computing won't just be faster — it will be increasingly AI-native, with more intelligence happening directly on our own devices.
And with the M6 and M5 Ultra, Apple is giving that vision considerably more processing power.