Are Apple’s new M5 Ultra and M6 its most powerful chips ever built?

Apple has introduced two new processors, the M5 Ultra and M6, as the company expands the performance and AI capabilities of its Mac lineup.

The M5 Ultra powers the new Mac Studio and is designed for demanding workloads, including 3D rendering, visual effects, and advanced AI models. The M6, built on a 2nm process, is aimed at a broader group of Mac users and developers running AI workloads locally.

Apple says the M5 Ultra is its most powerful chip to date, while the M6 adds more CPU and GPU cores, along with a redesigned Neural Engine.

M5 Ultra pushes Apple’s chip architecture further

Apple describes the M5 Ultra as its most powerful chip ever. The processor combines two dual-die M5 Max chips, creating the company’s first quad-die chip.

The M5 Ultra can be configured with up to 36 CPU cores and 80 GPU cores. It also provides 1.2 TB/s of unified memory bandwidth, which Apple says is 50% higher than the M3 Ultra introduced roughly a year and a half earlier.

The chip is designed for workloads that require substantial computing resources. Apple specifically positions the M5 Ultra for professional applications such as complex 3D rendering, visual effects, and running advanced AI models.

Its capabilities could also make the Mac Studio attractive to developers and hobbyists working with AI. The high-performance hardware provides additional capacity for users experimenting with demanding models without relying entirely on cloud computing.

M6 brings 2nm technology to the Mac

The M6 takes a different approach, targeting everyday Mac users while adding hardware designed for increasingly demanding AI workloads. Apple built the chip using a 2nm process and added a new CPU complex, additional CPU and GPU cores, and a Dual 16-core Neural Engine.

The M6 has a 12-core CPU complex, two more CPU cores than the M5. Apple says the chip can deliver nearly a 30% increase in peak GPU compute for AI compared with the M5.

That performance improvement is particularly relevant to on-device AI applications. Faster local processing can reduce the time needed to handle prompts when users run large language models directly on their Macs.

Apple also says the M6 provides greater unified memory bandwidth and improved energy efficiency, allowing the processor to handle demanding workloads while maintaining the advantages of local computing.

Apple targets local AI development

AI is becoming a central part of Apple’s strategy for its desktop processors. Rather than relying exclusively on cloud-based processing, the company is giving developers hardware and software tools to run and fine-tune AI models directly on Mac computers.

Apple says its developer frameworks automatically optimize workloads across the CPU, GPU, and Neural Engine. Developers can use Apple Foundation Models, App Intents, and their own proprietary AI models to build and run powerful AI workloads entirely on the device.

Apple’s broader AI architecture combines its own Apple Foundation Models with technology developed in collaboration with Google. Apple says its third-generation Foundation Models were custom-built in collaboration with Google and its Gemini technology, and these models help power the new Siri AI experience.

Some Siri AI processing can also use Apple’s Private Cloud Compute rather than running entirely on the device.

Apple has also built a long track record with local processing for sensitive features such as Face ID and Touch ID. Apple’s biometric architecture processes and protects Face ID and Touch ID template data through the Secure Enclave, and the biometric data does not leave the device.

That emphasis on local and privacy-focused processing could appeal to developers and users who want to experiment with AI while keeping more computation and data on their own devices.

M5 Ultra and M6 serve different users

The two processors are not direct replacements for each other. The M5 Ultra is built for maximum performance and is available in the Mac Studio, while the M6 is designed for the new Mac mini and is intended for a broader range of users.

The M5 Ultra supports configurations with up to 36 CPU cores and 80 GPU cores, along with 1.2TB/s of unified memory bandwidth. The standard M5 Ultra configuration in Mac Studio has a 30-core CPU and 64-core GPU, while higher configurations reach the 36-core CPU and 80-core GPU maximums.

Its architecture is designed for professional workloads that require substantial processing capacity and memory bandwidth.

The M6 instead focuses on bringing stronger AI performance to a smaller desktop. Its 2nm manufacturing process, additional CPU and GPU cores, and Dual 16-core Neural Engine are designed to improve everyday computing and local AI workloads.

That distinction allows Apple to address different parts of the desktop market, ranging from everyday users, developers, and AI hobbyists to professionals working with highly demanding applications.

New Mac mini starts at $899

The new Mac mini with M6 is available for preorder starting at $899. Apple says it will begin arriving at customers and become available at Apple Store locations and Apple Authorized Resellers starting September 22.

The starting price is higher than in previous Mac mini generations, but the M6 introduces significant changes to the processor itself. The move to a 2nm process and increased AI capabilities are central to the new model.

For users interested in running AI models locally, the M6’s additional GPU compute and expanded Neural Engine are particularly important. Apple designed the chip to improve on-device AI performance while also advancing power efficiency.

The M5 Ultra, meanwhile, prioritizes maximum performance and is aimed at users whose workloads justify the more powerful Mac Studio platform.

Apple’s chip strategy is increasingly tied to AI

The introduction of the M5 Ultra and M6 shows how AI has become a major consideration in Apple’s silicon development. The M5 Ultra focuses on raw computing capacity for professional AI and graphics workloads, while the M6 brings more specialized AI processing to a smaller desktop.

Apple’s strongest opportunity may be its ability to combine dedicated hardware with software frameworks that keep AI workloads on the device. That strategy could appeal to developers seeking to develop local models and users who prioritize keeping sensitive processing away from the cloud.

The M5 Ultra also establishes a new performance ceiling for Apple’s desktop lineup, with up to 36 CPU cores, 80 GPU cores, and 1.2 TB/s of memory bandwidth.

Whether these processors ultimately deliver Apple’s claimed performance in real-world workloads will depend on how they perform across applications. For now, the specifications show Apple pushing both its high-end workstation hardware and mainstream desktop silicon further toward AI-focused computing.

TL;DR

• Apple has introduced the M5 Ultra and M6 chips for its latest Mac Studio and Mac mini models.
• The M5 Ultra combines two dual-die M5 Max chips into Apple’s first quad-die processor.
• The M5 Ultra offers up to 36 CPU cores, 80 GPU cores, and 1.2TB/s of unified memory bandwidth.
• Apple calls the M5 Ultra its most powerful chip ever.
• The M6 is built using a 2nm process and has a 12-core CPU complex.
• Apple says the M6 provides nearly 30% higher peak GPU compute for AI than the M5.
• Apple is emphasizing local AI development, allowing developers to run and fine-tune models directly on Mac.
• The new M6 Mac mini starts at $899 and is available for preorder.
• Apple says the new Mac mini will begin arriving at customers and become available in stores starting September 22.

Don’t forget to follow us for more exclusive content.

If you liked this, you might also like:

• OpenAI warns ChatGPT’s Computer History could create new risks
• Gen Z is entering the workforce with a growing fear of AI replacement
• Walmart is finally letting shoppers use Apple Pay and tap-to-pay at checkout

This article was made with AI assistance and human editing.