An iPhone chip in a Mac: what Apple got right and where 8 gigs hurt
I was scrolling through the keynote live blog on my M1 Pro when the price appeared: $599. For a Mac. I went back up to make sure I hadn’t misread it.
The MacBook Neo isn’t my machine — I need the RAM and I need the ports — but what Apple did here deserves a closer look than “cheap Mac, neat.”
The bold part — an iPhone chip in a Mac
Here’s what happened: Apple took the A18 Pro — the exact chip inside the iPhone 16 Pro — and put it in a laptop running macOS. Six CPU cores (two performance, four efficiency), five GPU cores, fanless passive cooling. No M-series silicon anywhere. That’s a first.
The benchmarks tell a surprising story. According to early Geekbench 6 results, single-core performance hits 3,461 — that’s 47% above the M1’s 2,346. Multi-core lands at 8,668, roughly on par with the M1 and just below the M2’s 9,644. A phone chip matching a two-year-old laptop chip in sustained workloads, while running silent and starting at $599. I’ve been on Apple Silicon since the M1 Pro shipped — watching a phone chip match that machine’s single-core score is genuinely surreal.
But benchmarks aren’t the point. The point is that Apple just proved their chip architecture scales in both directions. The same silicon that runs your phone — the chip designed for a device you carry in your pocket — can run a full desktop operating system with a 13-inch Liquid Retina display attached to it. That’s the part worth paying attention to.
And it gets more interesting when you zoom out. There are credible rumours about a MacBook Ultra coming late 2026 — an OLED touchscreen Mac with an M6 chip, described as “touch-friendly, not touch-first.” Back in 2010, Steve Jobs dismissed the idea outright — “touch surfaces don’t want to be vertical,” he said at a press event. Now Apple is reshaping the Mac from both ends: an iPhone chip pushing up from below, touch pushing down from above. It’s hard to overstate how much the company has shifted.
The catch — 8 gigs, no way out
Here’s where it gets complicated. The MacBook Neo ships with 8GB of RAM. Not because Apple wanted to cut costs — because the A18 Pro was designed for 8GB. It’s a hardware limitation of the chip itself. There’s no 16GB option. There’s no upgrade path. You get 8 gigs, and that’s it.
For browsing, writing, email, and basic productivity — that’s fine. macOS has always been efficient with memory management, and 8GB will handle those workloads without drama. Apple Intelligence needs 8GB minimum to run, and you get exactly that — so Siri, writing tools, and the cloud-based AI features all work.
But here’s where it matters to me personally. I run Ollama with local language models on my M1 Pro. Even with 32GB, larger models push against the ceiling. Quantised 7B models run well; anything bigger gets tight. On 8GB, local inference is effectively off the table. You’re not running a 7B model alongside macOS and a browser — there’s simply no headroom.
What Apple has created, whether intentionally or not, is a two-tier AI strategy. Neo users get Siri, Apple Intelligence, and whatever cloud AI Apple provides. Pro and higher users get all of that plus the ability to run models locally — to experiment, to keep data on-device, to use AI without an internet connection. That’s a meaningful line, and 8GB is where Apple drew it.
Who is this actually for?
The Neo makes a lot of sense for specific people. Switchers from Windows or Chromebooks who want a real Mac experience without a $1,299 entry price. Students who need a laptop for notes, research, and light creative work. Anyone whose workflow lives in a browser and a text editor.
It’s not for developers. It’s not for anyone who works with large files, multiple VMs, or local AI tools. And that’s perfectly fine — Apple has the Air and the Pro for those people. The honest limitation is that the Neo won’t grow with you. If your needs expand beyond browsing and writing, you’ll hit the 8GB wall, and there’s no way through it.
What Apple does next
The MacBook Neo is Apple’s boldest hardware bet in years. The 8GB ceiling is the cost of that bet — and whether it pays off depends on what the next generation of phone chips can do.
The chip scales. The RAM doesn’t.