Apple Might Skip the M6 Entirely. Here's Why That Says More About AI Than Macs

Something odd is showing up in the Apple chip roadmap. The talk is that Apple could skip the M6 generation entirely and jump to an M7 built around AI. If that holds, it’s not a routine generational bump. It’s a signal that Apple is rethinking who its chips are actually for.

Let me be upfront about the sourcing first. This one hasn’t matured into a heavy community debate yet. There’s very little confirmed discussion floating around — this is the rumor-and-observation stage, not the leaked-spec stage. So rather than treat an unverified whisper as fact, I want to walk through why this pivot would make strategic sense, and what it would mean if it turns out to be real.

The M series was always a Mac chip

Start with the basics. The M series is the in-house silicon Apple rolled out in 2020 when it walked away from Intel. M1, M2, M3, M4 — these have been the brains of the MacBook, the Mac mini, and the Mac Studio.

The design goal was clear: performance per watt. Long battery life paired with enough muscle to chew through video edits and compiles without breaking a sweat. That was the M series identity.

Which is exactly why the word “skip” stands out. Apple rarely jumps over a whole generation. This is a company that has cranked the number up like clockwork, yearly or every other year. So the mere idea of skipping M6 is loaded.

Why “AI-focused” is the operative phrase

The center of gravity here is one component that keeps getting bigger inside the die: the NPU, or Neural Processing Unit. Think of it as the department that handles nothing but AI math.

For years, the CPU and GPU were the stars. The NPU was a supporting player, quietly helping with photo cleanup and voice recognition. Then generative AI arrived and changed the math. On-device AI — running models locally instead of phoning home to a server — became the competitive battleground.

Whether it’s a chatbot or image generation, keeping the work on the device instead of shipping it to the cloud leans hard on the NPU. It also demands far more memory bandwidth. So “AI-focused M7” really translates to one thing: shifting the design’s weight away from CPU and GPU and toward NPU and memory.

“Apple is abandoning Mac users” is the wrong read

Here’s the misconception worth clearing up. Going all-in on AI does not mean giving up on Mac performance.

The reason is simple: the Mac is Apple’s best stage for showing off AI in the first place. Running a large model locally takes a strong chip and generous unified memory, and the product with the most room for both is a desktop like the Mac Studio. There’s already a real trend of developers buying high-memory Macs specifically as local AI workstations — it’s a quiet but growing reason people pick the loaded configs over the base ones.

So the high-end line — M7 Pro, Max, Ultra — probably doesn’t vanish. It gets redefined as “the Mac that runs AI well.” This reads less like abandoning Mac users and more like handing the Mac a new job.

So what’s the actual play?

I read this pivot as Apple’s answer to two pressures.

The first is competition. Qualcomm, Intel, and AMD are all shouting “AI PC” and putting NPU numbers front and center in their marketing. The yardstick for what makes a good chip is shifting. In that climate, Apple showing up early with an “AI chip” business card is a natural move, not a surprising one.

The second is direction. Skipping a generation is a way of saying that nudging the existing roadmap up one notch isn’t enough. Leaving a number blank can be the point — a visible way to signal that the design philosophy itself has changed.

That said, worth repeating: this is still observation and interpretation. An M6 skip has not been confirmed in any shipping product, and that caveat matters.

The takeaway

Boil it down and it’s this. “Skip M6, go AI-first M7” is shorthand for two shifts at once: the chip’s protagonist moving from CPU to NPU, and the chip’s purpose moving from “fast Mac” to “smart AI machine.”

Which leaves one question. Next time you buy a Mac, will you stop asking “how many cores, how many gigs” and start asking “which AI models can I run locally”? If the answer is yes, then skipping a single chip number isn’t a small thing. It’s the opening shot in changing how we judge a computer at all.