Apple today has introduced the M5, the newest addition to Apple silicon chipsets, marking what the company calls its “next big leap” in AI and graphics performance. Built on third-generation 3nm process, the chipset has made its debut in the new 14‑inch MacBook Pro, iPad Pro, and Apple Vision Pro.
According to the tech giant, the new GPU architecture and Neural Engine (this is what Apple calls its in-house NPU) together claim to deliver some big performance improvements that are said to especially boost local AI processing throughput.
In terms of specs, the M5 packs a 10‑core CPU, with up to four performance cores and six efficiency cores. Apple says this design delivers up to 15 percent faster multithreaded performance compared to the M4. The GPU, also 10 cores, integrates a Neural Accelerator in each core, a first for Apple silicon. Due to these changes, Apple is promising over 4x peak GPU compute compared to M4, and more than 6x compared to the original M1 for AI‑specific tasks.
The Neural Engine itself remains 16 cores but is faster and more efficient, implying significant potential IPC (instructions per cycle) gains gen-on-gen.
Graphics are also said to see a very big bump up. The M5 introduces Apple’s third‑generation ray‑tracing engine, combined with "enhanced" shader cores and second‑generation dynamic caching. Together, these claim to deliver up to 45 percent higher graphics performance than M4 and as much as 2.5x faster than M1. Overall, Apple is promising smooth gameplay in graphically-heavy titles such as Cyberpunk 2077 on Mac, along with faster 3D rendering and video editing.
All the performance gains are not possible without sufficient memory bandwidth, and in this regard, Apple says it is delivering a near 30 percent boost at 153GB/s, and the capacity up to 32GB of unified memory.