Nate Silver, writing at his Silver Bulletin:

Last night, as President Trump delivered his State of the Union address, the Wall Street Journal reported that ABC News would lay off the remaining staff at 538 as part of broader cuts within corporate parent Disney. Having been through several rounds of this before, including two years ago when the staff was cut by more than half and my tenure expired too, I know it’s a brutal process for everyone involved. It’s also tough being in a business while having a constant anvil over your head, as we had in pretty much every odd-numbered (non-election) year from 2017 onward at 538/FiveThirtyEight. I don’t know all of the staffers from the most recent iteration of the site, but the ones I have met or who I overlapped with are all extremely conscientious and hard-working people and were often forced to work double-duty as jobs were cut but frequently not replaced. My heart goes out to them, and I’m happy to provide recommendations for people I worked with there.

Beyond that, I wasn’t inclined to say too much more, but it felt weirder not to say anything at all. And it’s easier to say something here than filter it through a reporter or something.

Jason Snell, at Six Colors:

With the M4 Mac mini being powerfully tempting for desktop Mac users who crave power, Apple has upgraded the Mac Studio to blast past the mini in terms of performance. The base model, still starting at $1999, is powered by the M4 Max chip previously available only in the M4 MacBook Pro. And the new high-end Mac Studio, starting at the same $3999 price tag, is powered by a monstrous chip with 32 CPU cores (including 24 performance cores) and up to 80 GPU cores. It’s a chip never seen before anywhere — the M3 Ultra.

You heard me. For Apple’s fastest Mac ever — and it’s clear that it will be — Apple’s shipping a chip based on two high-end chips (fused together with Apple’s UltraFusion technology) from Apple’s previous processor generation. Weird, right? It seems like a few things are going on here: first, that the development of the Ultra chip takes longer and that Apple won’t commit to shipping an Ultra chip in every chip generation. Second, that the first-generation three-nanometer chip process of Apple’s chipmaking partner, TSMC, isn’t as dead and buried as generally thought. Just this week Apple also introduced an iPad Air with an M3 processor, and of course the new iPad mini shipped with an A17 Pro processor based on the same process.

This M3/M4 generational fork — the M3 Ultra chip debuting in new Mac Studio models alongside the M4 Max — was so unexpected that, during my embargoed press briefing about the news yesterday, I thought the Apple rep misspoke when he said M3, not M4, for the Ultra models. But no, the Ultra chip really is a generation behind. When asked the obvious question — why — Apple’s answer was straightforward: the Ultra chips take a lot longer to engineer.

The M4 Max Studio models are, computationally, equivalent (exactly, I think) with the M4 Max MacBook Pros that debuted October 30, maxing out (no pun intended) at 16 CPU cores, 40 GPU cores, 8 TB of storage, and 128 GB of RAM. The M4 Max Studio models start at $2,000, but that starting price only gets you a 14-core CPU, 32-core GPU, 16-core Neural Engine, 512 GB of storage, and a measly 36 GB of RAM.

The intriguing M3 Ultra models start at $4,000, which gets you a 28-core CPU, 60-core GPU, 32-core Neural Engine, 1 TB of storage, and a healthy 96 GB of RAM. Available upgrades to, uh, ultra out the Ultra models:

• 32-core CPU + 80-core GPU: +$1,500
• 256 GB of RAM: +$1,600
• 512 GB of RAM: +$4,000 (only available with the 32-core CPU/80-core GPU)

SSD storage options for the Ultra models go up to 16 TB (a cool $4,600 over the base storage).

Jason Snell, writing at Six Colors:

Let’s start with the surprises. Both M4 MacBook Air models are priced $100 less than their predecessors: $1199 for the 15-inch model and $999 for the 13-incher. If I’m not mistaken, this is the first time that the new-generation design of MacBook Air introduced with the M2 chip has been available at the classic $999 price at launch. (The M1 Air, based on the Intel-era visual design, debuted at $999, but the M2 Air debuted at $1199 and only reached $999 when it was offered as an older model alongside the M3 Air.) As of now, the M4 Air can hold down the sub-$1000 price point all on its own, and previous models are mostly discontinued.

Another surprise is the the new color option: Space Gray is out. The ultra-dark-blue Midnight remains, as do the classic Silver and hint-of-champagne Starlight. The new color is Sky Blue, which apparently is a metallic light blue that really shows itself as a color gradient when viewed at various angles.

Very cool that the new M4 starts at $999. Each successive generation of Apple Silicon, at least in laptops, is getting more and more predictably regular.

Mark Gurman, reporting for Bloomberg on 6 December 2024, regarding Apple’s first in-house cellular modem, “Apple Plans Three-Year Modem Rollout in Bid to Top Qualcomm”:

For now, the modem won’t be used in Apple’s higher-end products. It’s set to come to a new mid-tier iPhone later next year, code-named D23, that features a far-thinner design than current models. The chip will also start rolling out as early as 2025 in Apple’s lower-end iPads.

We now know the name of that modem, the C1, from its debut in the iPhone 16e last month. Then, also on December 6, in a separate report headlined “Apple Explores Macs, Headsets With Built-In Cellular Data”:

The first modem will also appear in low-end iPads next year, with the 2026 update coming to Pro versions of the iPhone and iPad.

The cellular models of the new 11th generation iPads announced yesterday do not, it turns out, use the C1. The specs don’t match those of the iPhone 16e, and when I asked an Apple representative, they confirmed that none of the new iPads (including the Airs) use the C1 modem. (But, Apple reassured me, they all offer terrific cellular networking.)

I’m not saying Gurman was wrong, because there are nine full months left in 2025 for Apple to release a 12th-generation low-end iPad with the C1. The previous (10th) generation came out in October 2022, but the 9th generation came in September 2021, just 13 months prior. And this week’s new M3 iPad Airs replaced M2 models that arrived just 10 months ago. But, you know, it sure seems doubtful Apple is going to rev this hardware in 2025, so I’ll place my bet that he was wrong about this too.

(And yes, a Bloomberg Terminal subscription really does start at $32,000/year per seat.)

Mark Gurman, in his Power On column for Bloomberg, on January 12:

The new entry-level iPads — J481 and J482 — will get faster processors and Apple Intelligence. The current models have the A14 chip and 4 gigabytes of memory. Look for the new versions to have the A17 Pro chip, matching the iPad mini, and a bump to 8 gigabytes of memory. That’s the minimum needed to support the new AI platform.

The new iPads sport the A16 chip and thus do not support Apple Intelligence. But who cares about little details like that when you know the codenames, which is what really matters.

I’ll bet what happened is that Gurman was right, and the new iPads were set to use the A17 Pro chip and support Apple Intelligence. But after Gurman spoiled it seven weeks ago, Apple scrapped those plans and changed the chips to the A16 just to spite him.