tenthousandthings

For me, it is all about the display, as a visually-oriented person who works in the museum and art world. So Pro it is.

Most of my actual work with images and color takes place on macOS, but I have an old Pro I use for just looking. It’s had a good run, but this new display is a “must have” from my point of view. 

neilm said:

The key question is: what will be the device hardware requirements for AI aided services?

For instance my iPhone XS is the oldest that can run the current iOS 17, so it may well not be supported by iOS 18. While my XS still works well — even the original battery! — and does everything I want it to, I may be prodded to upgrade by the availability of useful AI.

I think the worst case is it will be A14/M1 and above. That’s when they introduced the 16-core Apple Neural Engine NPU. But maybe the cutoff will be an 8-core NPU, which was introduced with the A12/A12X, so your XS might be okay? One last hurrah before it gets left behind next year?

blastdoor said:

I recently bought a refurbished m2pro Mac mini because I gave up on an m3 mini being released. I also would not be surprised if the studio skips m3 too. 

With regard to the Mini and the iMac refresh cadences, whatever. It’s fine. But if Apple skips the M3 Ultra, they need to explain what they are doing with the silicon. It raises a lot of questions, not all of which are (how shall I put it?) strategic.

It’s not enough to just say it’s all product-driven, and they only build the silicon their products need. Everyone understands why Apple doesn’t talk about unreleased products. But expectations need to be managed.

Marketing and management needs to wake up and realize the architects and engineers need to be allowed to explain more than what is currently being explained.

nubus said:

godofbiscuitssf said:

Apple has been doing ML integrations for years.  Marketing-speak changed “ML” to “AI” and you’re buying into it.  

As for this article, Apple’s been updating the Neural Engine every time in the silicon, sometimes obviously with the number of cores, sometimes by stating so, with giant performance increases, and other times by changing the architecture of the core.  

Apple has indeed been delivering a Neural Processing Unit for years. However, even M3 Max is only delivering the ML performance of iPhone 14 Pro (17-18 TOPS). Intel is doing 34 TOPS with their 2023 laptop CPUs (combined NPU+GPU) and Lunar Lake will launch this year with 100 TOPS performance for laptops. Standing still is why the stock is dropping.

Something will need to change. I would have killed M3 Ultra and asked the entire team to work on M4 or refocused M3 Ultra on ML at any cost to show at WWDC.

The presentation on this at WWDC will be interesting. I can't find it right now, but I read somewhere there is no difference between the NPU in the A17 Pro and the M3, and the 35 trillion versus 18 trillion numbers reflect two different standards for measuring it. So one number is for the mobile industry, the other for the computer industry. Apple didn't double the peak NPU ops/sec in the A17 Pro versus the A16 Bionic. Instead, the competition changed to a different standard that provided better-sounding numbers, and Apple was forced to adopt that for the A17 Pro and the iPhone 15 Pro without comment, but they kept the old standard for the M3, to avoid appearing to claim an astounding leap forward in the NPU relative to the M1 and M2, when in reality it is incremental.

I'll try to find it, it wasn't an anonymous internet commenter or anything, it described the change, but it was only like two sentences and now I can't find it.

Gurman is betting that A18/M4 is on N3E, which has been in volume production since late last year, so why not? There are a lot of unspoken "ifs" in that prediction, however. The first being "If M4 is on N3E..." 

It's fun to see the code names, and I think Apple Insider is right to guess they are for M4, M4 Pro, M4 Max, but they don't tell us anything beyond confirming (if that guess is correct) the M4 launch will have the same marketing/naming structure as the M3 launch.