Apple’s new M5 chipset implements a new GPU architecture where each core features Neural Accelerators.
The M5 SoC also brings a unified memory bandwidth of 153GBps
Click Here to Add Gadgets360 As A Trusted Source
Apple launched new hardware, a refreshed 14-inch MacBook Pro, Vision Pro, and a new iPad Pro on Wednesday. All of these were equipped with the Cupertino-based tech giant’s new M5 chipset, featuring up to a 10-core CPU and 10-core GPU architecture. The company says the M5 system-on-chip (SoC) was designed keeping artificial intelligence (AI) workflows in mind. It is said to offer four times the peak GPU performance as compared to the previous generation. Notably, each GPU is also equipped with Neural Accelerators to optimise the workloads.
Breaking Down Apple’s M5 Chipset
In a newsroom post, the tech giant announced and detailed the latest Apple silicon. It is built on the third-generation 3nm process technology and introduces a new GPU architecture. The M5 chip features a 10-core GPU architecture, where each core is equipped with a Neural Accelerator. Apple says this enables the chipset to power heavier AI workloads, such as “running diffusion models in apps like Draw Things, or running large language models locally using platforms like webAI.”
Compared to the M4 chipset, the company claims the latest SoC achieves 30 percent faster performance. It also brings Apple’s third-generation ray-tracing engine, which delivers up to a 45 per cent graphics uplift in apps compared to the previous generation. For end users, it means smoother gameplay, more realistic visuals while running 3D apps, and faster rendering times for complex graphics projects.
Coming to the AI workflow, the M5 chip brings the same 16-core Neural Engine, which the company claims is faster than the last generation. Apart from optimising performance when users work on AI models, the Neural Engine also improves the performance of Apple Intelligence tools. This means apps such as Image Playground will now be more responsive and generate output faster. Additionally, any third-party apps that are built on the company’s Foundation Models framework will also witness faster performance.
Finally, the M5 SoC also brings a unified memory bandwidth of 153GBps, which is claimed to be nearly 30 percent higher than its predecessor. On one hand, it boosts the overall capability of the CPU, GPU, and Neural Engine, and on the other, it also delivers higher multithreaded performance in apps and games. Additionally, the chipset supports a memory capacity of 32GB.
