Apple’s New Mac Mini and Mac Studio Bring AI Computing to the Desktop

Apple has begun selling its latest Mac mini and Mac Studio computers, putting greater emphasis on artificial intelligence workloads that can be processed directly on desktop hardware.

The new range includes the Mac mini with M6 and M5 Pro chips and the Mac Studio with M5 Max and M5 Ultra. Apple is highlighting their performance for AI, professional applications and other demanding workloads. The company is also promoting local processing as an alternative to relying entirely on cloud-based computing.

New Macs Bring Different Levels of AI Performance

The two desktop families target different levels of computing requirements.

The Mac mini is the smaller and less expensive option, while the Mac Studio is designed for professional workloads that require substantially more processing power and memory.

Model	Processor	Starting Price in the U.S.	Maximum Unified Memory
Mac mini	M6	$899	32GB
Mac mini	M5 Pro	$1,699	Higher than M6 configuration
Mac Studio	M5 Max	$2,499	128GB
Mac Studio	M5 Ultra	$5,499	512GB

Apple lists these as starting U.S. prices, while configuration and availability can vary by market.

Mac Mini Gets the New M6 Chip

The M6 is the main addition to the new Mac mini. Apple says it is built using a 2-nanometre manufacturing process and includes a 12-core CPU, 12-core GPU and Dual 16-core Neural Engine.

Apple reports that the M6 Mac mini can deliver up to four times the AI performance of the previous M4 model. The company also says it can provide up to 40% faster CPU performance compared with the previous generation. These figures are based on Apple’s stated testing, so they should not be treated as universal performance results for every application or AI model.

The M6 model can also be configured with up to 32GB of unified memory, while its memory bandwidth reaches up to 170GB/s.

Mac Studio Is Built for Larger Workloads

The Mac Studio occupies the higher end of Apple’s desktop range.

Its M5 Max version starts at $2,499, while the M5 Ultra version begins at $5,499. The M5 Ultra configuration can provide up to 36 CPU cores, 80 GPU cores and 512GB of unified memory, with memory bandwidth of up to 1.2TB/s.

The large memory capacity is particularly relevant to AI applications that require substantial resources. Apple says the M5 Ultra Mac Studio can run very large language models entirely on the device.

That capability does not mean every large AI model will run at the same speed. Performance can vary according to the model, software implementation and workload.

Multiple Mac Studios Can Work Together

Apple is also using its latest connectivity technology to demonstrate how several Mac Studios can be combined.

The company has shown multiple Mac Studio systems working together on demanding AI workloads. Reuters reported an Apple demonstration involving four Mac Studios running a trillion-parameter AI model and working on a graphics coding task.

The computers can communicate through Thunderbolt 5 using Remote Direct Memory Access, or RDMA. Apple says this technology can allow multiple Mac Studio systems to be clustered for distributed AI inference, with the company reporting up to three times the performance of a single system in its stated comparison.

These figures describe Apple’s demonstrations and testing rather than a guarantee for all distributed AI workloads.

Unified Memory Plays a Key Role

Unified memory remains an important part of Apple’s approach to high-performance computing.

Apple Silicon allows different processing components to work with a shared memory pool rather than relying on the same type of separate CPU and GPU memory arrangement found in many conventional systems.

For AI applications, having a large shared memory pool can be useful when working with models that require significant amounts of memory. The M5 Ultra Mac Studio’s support for up to 512GB gives professional users considerable capacity for local workloads.

Apple Is Targeting Cloud-Based AI Workloads

Apple’s desktop launch comes as businesses and developers evaluate different ways of running AI applications.

Reuters reported that Apple is presenting its new Macs as an alternative to cloud-based AI services, particularly for organisations that want to process workloads locally. Apple’s argument is that purchasing hardware can reduce reliance on usage-based cloud computing costs for certain workloads.

However, whether local hardware is more economical depends on factors such as the workload, hardware configuration, electricity costs, software requirements and the amount of computing needed.

Competition Extends Beyond Apple

Apple is entering a market where Microsoft and Nvidia already have significant positions.

Reuters, citing IDC analyst Linn Huang, reported that Apple has about 4.6% of the enterprise desktop market, compared with 91.3% for Windows. Microsoft is also pursuing local AI processing through Windows and related developer tools, while Nvidia remains a major supplier of AI computing hardware for data centres.

Apple’s new desktop strategy therefore represents an expansion of its AI computing approach rather than a replacement for the wider cloud and data-centre ecosystem.

New Mac Mini and Mac Studio Are Available

Apple says the new Mac mini and Mac Studio are now available through Apple Store locations, its online store and the Apple Store app. The M6 Mac mini starts at $899, while Mac Studio pricing begins at $2,499 for M5 Max and $5,499 for M5 Ultra.

With these systems, Apple is combining its latest processors, large unified-memory configurations and Thunderbolt 5 connectivity to support increasingly demanding AI and professional workloads.

FAQs

What is the starting price of the new Mac mini?

The M6 Mac mini starts at $899 in the U.S., while the M5 Pro version starts at $1,699.

How much memory does the M5 Ultra Mac Studio support?

The M5 Ultra Mac Studio supports up to 512GB of unified memory.

Can the new Macs run AI models locally?

Yes. Apple specifically promotes the new Mac mini and Mac Studio for on-device AI workloads and says the M5 Ultra Mac Studio can run very large language models locally.

What is RDMA over Thunderbolt 5 used for?

Apple uses RDMA over Thunderbolt 5 to connect multiple Mac Studio systems for distributed workloads, including AI inference.

Is local AI always cheaper than cloud computing?

Not necessarily. The financial impact depends on the type and volume of workloads, hardware costs, electricity, software and other operational factors.

Conclusion

Apple’s latest Mac mini and Mac Studio place local AI computing at the centre of the company’s newest desktop strategy. The M6 brings increased AI capabilities to the compact Mac mini, while the M5 Ultra Mac Studio offers much larger memory capacity for demanding professional and AI workloads.

Apple is also demonstrating how multiple Mac Studios can be connected for larger workloads. Rather than eliminating the need for cloud or data-centre computing, the new Macs give developers and businesses another option for processing AI tasks directly on desktop hardware.