A 5nm, a14 or a15 based MacPro is suicide for the MacPro. But realistically, there’s not going to be an Apple silicon MacPro. The volume is musicale, and the hardware expectations are a galaxy away from what the M series can deliver. It’s just smarter to let that line finally die. Either that or we get a Max Studio with a couple non-gpu PCIe slots for $6k and up, and I think we can all agree just letting the name die is better than that.
It’s not about matching with AMD or intel, it’s about spending a lot of money and effort to be the launch customer for 3nm and then letting someone else get 3nm into computers first. The 5nm roll out may have been slow, but AMD only got to 5nm in late 2022, a year after the M1pro. Meanwhile, AMD’s roadmap still calls for 3nm Zen5 and Zen5c to launch during 2024. Apple should be concerned about zen 5, since laptop zen 4 is already nipping at m1pro’s perf/watt when in low power modes. If AMD gets a node advantage, Apple will lose their Perf/watt advantage, and I don’t see them willing to do that.
Also, realistically, even 1 million m3max dies likely takes less than 10k wafers (assuming the die is even larger than the M1 Max despite the denser node). Meanwhile TSMC was making almost 100k 5nm wafers per month by the end of 2020. Assuming a similar ramp up, the entire first quarter of MBP sales can be produced from 3 days worth of wafers. That’s inconsequential to iPhone production, especially since Apple can easily launch non-pro phones with the a16 to help yields for the a17.
But we can speculate all day. My main point is that there seems to be only a vanishingly small chance that an m2pro ever launches. The hardware and designs in the a15/m2 are far too out of date and the vastly better a17 and n3 node too close to release to make an m2pro feasible. Even if we get a winter/spring MBP on n5P or n4, it has to be based on a16 and therefore would still be named m3 pro not m2 pro. If the M2pro ever existed (and I don’t think it did, I always assumed the bigger, low volume dies would skips generations), it doesn’t now. Or we’d already have seen it.
That’s a lot of big words to say apples going to switch everything to 3 nm in six months.
NOT GOING TO HAPPEN!
To start with, again, Apple does not care about having the first 3 nm computers on the market.
It won’t affect their bottom line, and they still make products to this day still based off of 7NM.
As for the speculation that Apple might discontinue the Mac Pro, it’s nonsense. They’ve already officially announced an Apple Silicon Mac Pro “in name” is coming.
They announce that at the same event as the original Mac Studio.
Even if it’s not the Mac Pro *you* want, it’s still a Mac pro.
And again, Apple is not introducing processors, they are introducing products.
They don’t have to compete with AMD or Intel or have their processors out first, because they know most of their customers will not care.
The A12Z iPad Pro and the M2 Air prove this, both were introduced right before a new architecture went into production, Apple still released them anyway, and they still sold like hotcakes.
People aren’t going to switch to a PC just because Apple didn’t release 3 nm when they liked.
All the rumors from very reliable sources, and even leaked benchmarks all point to new MacBook pros on the 5 nm process, and the same is likely true with the Mac Pro.
These are my expectations, backed up by Apple’s previous scheduling and reliable sources.
Again, Apple couldn’t care less about how fast they come out with chips compared to the competition. If they did, they wouldn’t still be selling almost an entire desktop lime powered by M1.