Apple has officially unveiled its groundbreaking M5 chip, marking the company's most significant leap in AI performance for Apple Silicon to date. The new processor debuts in the 14-inch MacBook Pro, iPad Pro, and Apple Vision Pro, all available for pre-order today with availability beginning October 22.
The M5 represents a fundamental shift in Apple's chip architecture, introducing Neural Accelerators directly into each GPU core for the first time.
This innovative design delivers over 4x the peak GPU compute performance for AI workloads compared to the M4, and an impressive 6x improvement over the original M1 chip. Built on third-generation 3-nanometer technology, the M5 doesn't just excel at AI—it brings substantial improvements across every aspect of computing performance.
Revolutionary GPU Architecture Powers AI Renaissance
At the heart of the M5's AI prowess lies its next-generation 10-core GPU, each core equipped with a dedicated Neural Accelerator. This architectural innovation enables AI-driven workflows like diffusion models in apps such as Draw Things and large language models on platforms like webAI to run dramatically faster on the device.
The GPU also features enhanced shader cores and Apple's third-generation ray-tracing engine, delivering up to 45 percent better graphics performance in ray-traced applications and up to 30 percent faster graphics overall compared to M4.
The chip's 16-core Neural Engine complements the GPU's Neural Accelerators, creating a fully optimized AI ecosystem.
This combination accelerates Apple Intelligence features, including Image Playground and on-device AI models, while developers using Apple's Foundation Models framework will experience significant performance gains. The M5's unified memory architecture offers 153GB/s bandwidth—a nearly 30 percent increase over M4—enabling devices to run larger AI models entirely on device with 32GB memory capacity.
CPU Performance Reaches New Heights
The M5 features what Apple claims is the world's fastest performance core, with a 10-core CPU configuration that includes up to four performance cores and six efficiency cores.
This design delivers up to 15 percent faster multithreaded performance compared to M4, allowing professionals to seamlessly run demanding creative suites like Adobe Photoshop and Final Cut Pro simultaneously while uploading large files in the background.
For Apple Vision Pro users, the M5 brings tangible improvements beyond raw performance numbers. The headset now renders 10 percent more pixels on its micro-OLED displays with refresh rates increasing to 120Hz, resulting in crisper details, more fluid display performance, and reduced motion blur.
AI-powered features like transforming 2D photos into spatial scenes or generating Personas operate with greater speed and efficiency.
Developer-Friendly and Future-Ready
Applications built with Apple's frameworks—including Core ML, Metal Performance Shaders, and Metal 4—will automatically benefit from the M5's performance improvements. Developers can also directly program the Neural Accelerators using Tensor APIs in Metal 4, opening new possibilities for AI-enhanced applications.
The M5's power-efficient performance aligns with Apple's ambitious Apple 2030 carbon neutrality goals, helping the new devices meet stringent energy efficiency standards while reducing lifetime energy consumption.
As Apple continues its silicon revolution, the M5 establishes a new benchmark for on-device AI computing, demonstrating that desktop-class AI performance no longer requires compromises in battery life or thermal management.
Key Features of Apple M5 Chip
- Next-Generation GPU: 10-core GPU with Neural Accelerators in each core
- AI Performance: Over 4x peak GPU compute for AI vs. M4; 6x vs. M1
- CPU Configuration: Up to 10 cores (4 performance + 6 efficiency)
- CPU Performance: Up to 15% faster multithreaded performance vs. M4
- Neural Engine: Enhanced 16-core Neural Engine for on-device AI
- Graphics Performance: Up to 45% improvement in ray-traced applications
- Memory Bandwidth: 153GB/s unified memory (30% increase over M4)
- Memory Capacity: Up to 32GB unified memory
- Manufacturing Process: Third-generation 3-nanometer technology
- Ray Tracing: Third-generation hardware ray-tracing engine
- Apple Intelligence: Optimized for on-device AI models and features
M5 vs M4: Performance Comparison
| Feature | M5 | M4 | Performance Gain |
|---|---|---|---|
| GPU Cores | 10-core with Neural Accelerators | 10-core | 4x AI compute |
| CPU Cores | Up to 10 (4P + 6E) | Up to 10 (4P + 6E) | 15% faster multithreaded |
| Neural Engine | 16-core (enhanced) | 16-core | Faster AI inference |
| Graphics Performance | Up to 45% faster (ray tracing) | Baseline | 30-45% improvement |
| Memory Bandwidth | 153GB/s | 120GB/s | 27.5% increase |
| Memory Capacity | Up to 32GB | Up to 32GB | Same |
| Manufacturing | 3nm (3rd gen) | 3nm (2nd gen) | More efficient |
| AI Performance (GPU) | 6x vs. M1 | 1.5x vs. M1 | 4x improvement |
| Ray Tracing | 3rd generation | 2nd generation | Enhanced quality |
| Neural Accelerators | Integrated in GPU | Not available | New capability |
Note: Performance improvements are based on Apple's official specifications and may vary based on specific workloads and configurations.