Apple Unleashes M5 Chip Built On 3nm With Huge Uplifts In AI, CPU And GPU

Apple M5 render (closeup).
Apple introduced the next iteration of its custom silicon called M5 with the promise of massive performance uplifts in artificial intelligence, CPU workloads, and GPU chores. Built on a 3-nanometer manufacturing process, the Arm-based chip designed in-house comes with a series of big claims. One of those is Apple touting the "world's fastest performance core."

On the CPU side, the M5 wields 10 cores consisting for four performance cores and half a dozen efficiency cores. According to Apple, the CPU delivers 15% faster multi-threaded performance compared to the previous generation M4. Curiously, despite claims of having the fastest performance core on the planet, Apple's announcement spends the least amount of time focusing on the CPU, with a bigger emphasis placed on the GPU and onboard neural engine.

Perhaps that is because those two elements work in tandem to some extent. Apple says the M5's next-generation GPU architecture is optimized for both AI and graphics, with each of the 10 GPU cores having their own neural accelerator. According to Apple, this results in a 4x uplift in GPU compute compared to the M4, and over 6x peak GPU compute for AI performance versus the M1.

Die diagram of Apple's M5 chip.

The GPU also sports enhanced shader cores in the M5 for up to 30% faster performance when pitted against the M4 (and up to 2.5x faster performance than the M1). Meanwhile, the M5's third-generation ray tracing engine purportedly delivers up to a 45% graphics uplift in ray-traced applications.

"Combined with rearchitected second-generation dynamic caching, the GPU provides smoother gameplay, more realistic visuals in 3D applications, and faster rendering times for complex graphics projects and other visually intensive applications," Apple says.

Once again, Apple is sticking to 16 cores for its neural engine, though the company claims it's even faster this time around. Combined with the neural accelerators in the CPU and GPU cores, as well as an almost 30% increase in unified memory bandwidth, which now sits at 153GB/s, Apple says M5 is fully optimized for AI workloads.

"M5 ushers in the next big leap in AI performance for Apple silicon," said Johny Srouji, Apple’s senior vice president of Hardware Technologies. "With the introduction of neural accelerators in the GPU, M5 delivers a huge boost to AI workloads. Combined with a big increase in graphics performance, the world’s fastest CPU core, a faster neural engine, and even higher unified memory bandwidth, M5 brings far more performance and capabilities to MacBook Pro, iPad Pro, and Apple Vision Pro."

The first devices to receive the M5 upgrade include the 14-inch MacBook Pro, the iPad Pro, and a retooled Vision Pro.