Report: Apple to skip M6 Pro/Max chips, fast-track M7 for local AI

Apple’s M-series release cadence has been very predictable so far. According to a new report from Mark Gurman at Bloomberg, that’s about to change, and it’s all about getting more powerful products to market sooner.

Here’s how it works now. First, Apple launches a standard low-end M chip (M4, or M5, for example) that is found in high-volume products such as MacBook Air, Mac mini, and the entry-level MacBook Pro. That’s followed roughly six months later by higher-end Pro and Max versions with more CPU and GPU cores, for more expensive high-performance products. And finally, Apple occasionally ships the highest-end “Ultra” versions of the chip family (basically two Max chips stitched together), last seen in the M3 generation.

According to Gurman’s report, Apple’s base M6 is on track to launch as early as this fall in Apple’s entry-level products. However, he claims Apple will not debut the M6 Pro and Max variants in the spring as usual. Instead, the company is looking to fast-track the M7 generation, moving up its release dates by as much as half a year.

That would mean products with an M7 processor (code-named Delos or H19G) could arrive as early as the first half of 2027, when the M6 Pro/max would usually arrive. M7 Pro and M7 Max products (code-named H19S and H19C) would then debut in late 2027 with the M7 Ultra (H19D) arriving in 2028.

In other words, it would appear that we’re getting the M6 on time, but then the entire M7 generation is coming half a year early. Gurman also reports that an M5 Ultra (code-named H17D) is still expected this year in an updated Mac Studio.

The race to provide the best on-device AI processing is hotter than ever right now, and this appears to be the reason behind the shakeup. Nvidia made waves with the introduction of its RTX Spark chips coming to Windows laptops starting late this year, and AMD, Intel, and Qualcomm are all rushing to make the best chips for consumer AI as well.

The M7 generation is expected to bring a significant boost to on-device AI performance. The base M7 will have memory bandwidth of around 240 gigabytes per second, a 20 percent boost over the rumored 200 GB/sec of the coming M6, and about 57 percent higher than the 153 GB/sec of the current M5. Pro and Max versions scale bandwidth up from there.

Beyond memory bandwidth, both GPU performance and Neural Engine performance are crucial to on-device AI processing. The M6 is alleged to have a new-generation GPU with more cores, an improved Neural Engine, and better performance across all processing cores. It’s unclear what further advancements the M7 will have, but one should expect an AI-focused chip to focus on GPU and Neural Engine improvements.