Apple’s M4: A Big Step Forward or a Big Yawn?

Apple's new M4 chip focuses heavily on energy efficiency and AI enhancements, offering more efficient cores and a powerful Neural Engine.

Nick Bild
22 days agoMachine Learning & AI
The M4 SoC has just joined the M-series line of chips (📷: Apple)

Apple has just announced the latest addition to their line of custom M-series chips, the M4 system on a chip (SoC). The fact that this SoC is being introduced not in a desktop or laptop as usual, but rather in the new iPad Pro, says a lot about the direction that Apple is going with this technology. Sure, the performance is better than previous generations of M-series silicon, but it is nothing that will really knock your socks off. Rather, a heavy focus was placed on energy efficiency and artificial intelligence (AI).

As for those performance enhancements, the four performance cores of the M3 will remain the same in the M4, but there will also be a three-core version available to reduce costs and energy consumption. The number of efficiency cores will increase from four to six across the board, however, prioritizing more efficient compute options. A few enhancements were made to the onboard GPU as well, like adding dynamic caching, hardware acceleration for ray tracing, and mesh shading. There is nothing revolutionary about this, but it will be a significant improvement for iPads, anyway, which are currently equipped with M2 processors.

One area where the M4 really shines is in its focus on AI. The new Neural Engine can perform 38 trillion operations per second, which is more than twice what the M3’s Neural Engine is capable of. The chip also features machine learning accelerators in the CPU itself to further speed up the execution of AI workloads. Apple claims that this makes the M4 more powerful than the neural processing unit found in any AI-focused PC today, so this technology has the potential to enable the local execution of many useful, real-time AI applications.

Aside from being better for the environment, the energy efficiency of the M4 also has the potential to allow devices to operate for longer between recharges. That will be an especially nice feature for iPads and laptops, although no specific details were given on how this impacts battery life in the real world.

Unless you are excited for the AI processing power of the M4, this chip might not seem especially interesting based on the initial specs. However, it is not yet clear what the M4 might be able to do when it finally arrives in a laptop or desktop. The energy and thermal constraints of an iPad are considerably different than these devices, so the true potential may not be known just yet. With active cooling and a beefier power supply, the M4 might prove to be more impressive than we expect.

Nick Bild
R&D, creativity, and building the next big thing you never knew you wanted are my specialties.
Latest articles
Sponsored articles
Related articles
Latest articles
Read more
Related articles