Apple Announces M5 Chip with Over 4x Peak GPU Compute for AI Compared to M4

Apple announced the M5 system on a chip, built using third-generation 3-nanometer technology, for the 14-inch MacBook Pro, iPad Pro, and Apple Vision Pro. The devices are available for pre-order.

The M5 features a 10-core GPU architecture with a Neural Accelerator in each core, delivering over 4x the peak GPU compute performance for AI compared to M4 and over 6x compared to M1. Graphics performance reaches up to 45 percent higher than M4 with third-generation ray tracing and up to 30 percent faster than M4 with enhanced shader cores. Compared to M1, graphics performance is up to 2.5x faster.

The chip includes an up to 10-core CPU with six efficiency cores and up to four performance cores, providing up to 15 percent faster multithreaded performance over M4. A 16-core Neural Engine and unified memory bandwidth of 153GB/s represent a nearly 30 percent increase over M4 and more than 2x over M1. Memory capacity is 32GB.

The GPU integrates with Apple frameworks including Core ML, Metal Performance Shaders, and Metal 4. Developers can program Neural Accelerators using Tensor APIs in Metal 4. Applications such as Draw Things and LM Studio run diffusion models and large language models locally on the new 14-inch MacBook Pro and iPad Pro.

Apple Vision Pro with M5 renders 10 percent more pixels on micro-OLED displays at refresh rates up to 120Hz. AI features...