tht

The A16 Bionic had been rumored to be being developed on the 4nm process but was released on the 5nm process. This change seems to corroborate The Information's story, though it was only a rumor.

Apple touts that A16 Bionic is fabbed on TSMC 4nm. This sentence would make more sense if you used TSMC 3nm.

A full node change is essentially the Moore's law doubling of transistors, give or take. Chip vendors would typically have to change their designs to adapt to the new node, such as going from TSMC 5nm to 3nm. With half node changes, they can typically keep the design rules for 5nm, but enjoy something like a 10 to 20% increase in transistor count, 5 to 15% power reduction, or some combination. I suppose TSMC 4nm can be considered an enhanced TSMC 5nm, with better transistor density, but just call it TSMC 4nm.

Calling it 5nm makes it sound like there isn't improvement, and that's factually incorrect. It's a half node improvement that TSMC has done for basically a decade now.

If in 2021, Apple was expecting TSMC 3nm to be in mass production by summer of 2022, which would be in time for fall iPhone shipments, they would have designed a chip given TSMC 3nm capabilities, like hardware ray tracing features. Once TSMC and Apple saw that 3nm was going to make it on time, and they would have figured it out in 2021, they would fallback to the half node step, TSMC 4nm, and get the typical half node improvements, like 10 to 15% performance, 10 to 15% less power, some combination.

Strategically, I kind of think the Jade C die (M1 Max), Jade C chop (M1 Pro), Jade 2C (M1 Ultra), and the failed Jade 4C was a mistake. It didn't scale in the manner that buyers wanted. The M1 Max and on down appear fine. The M1 Ultra and on up? There have been issues. Apple's designed itself into a box that can't get them to ship higher end machines. That's a bigger issue than TSMC being late.

jeromec said:

I am a bit confused.
It seems to me that external display support has been here forever for iPads.

I believe the new thing is external display support with Stage Manager, which is a big deal, and could be called "full" external display support.

The prior external display feature was mirroring. Whatever was on your iPad display was mirrored on the external display. Specific apps could be designed to use the external display to extend the display area. Eg, a video editor could display video on the external display while showing its UI on the iPad. 

danox said:

tht said:

mikethemartian said:

tomowa said:

Just asking, as I am not a Tesla car user. 

Does Tesla sell subscriptions for "enhanced features".  

Does Tesla allow 3rd party software/firmware access to its car's operating systems? Or does Tesla control the whole environment?
Is there any irony in Musks position, regarding Apple here?

I’m not sure but I think they charge a one time fee to unlock features but I don’t think it is an ongoing subscription.

These are the current aftermarket software unlocks Tesla is offering for my Model 3.

One time items consisting of:
$2000 Acceleration Boost
$6000 Enhanced Autopilot
$15000 Full Self Driving

Subscriptions consisting of:
$199/mo Full Self Driving
$9.99/mo Premium Connectivity

Tesla's on-screen software platform is obviously Tesla only with contracts for games and services. I'm not sure if, say, they are getting a cut from Spotify subscriptions. They might be. It's not a free for all App Store. You get what you get from Tesla. I don't know what their cancellation policy is.

The software unlocks are at a per-owner level too I think, but perhaps that has changed. The software locks are linked to the account holders, not the car. So if you sell the car to someone else, they may not get the software unlocks and they would have to buy them. Not sure where that landed.

If you sell your model 3 on, does the new owner get to keep/use those features too?

No. The software unlock is per model per account AFAIK.

I buy the acceleration boost for $2k for my Model 3. Enjoy it for a few months. I then sell the car to someone else. That new owner will not have acceleration boost. Same policy with FSD.

If you trade in you car to Tesla or a dealer, the software unlocks will most certainly be erased, and new owners will have to by the software unlocks for that vehicle. It's per account per vehicle.

king editor the grate said:

How many processors can 600,000 wafers yield?

(Mmm. Wafers.)

TSMC uses 300mm diameter wafers, which is about 70000 square millimeters. An Apple phone SoC is about 90 sq mm. An Apple Mac SoC is about 120 sq mm. Using 100 sq mm per SoC, that equates to 700 SoC chips per wafer. Not all chips are functional coming off the wafer, and assuming yield percentages are on order 90% -Lower at the start, higher after a few years - there's about 600 usable chips at 100 sq mm.

600,000 wafers, presumably per year here, is 360 million SoC chips at 100 sq mm. Quantities get exponential worse as the chips get bigger. Ie, a 200 sq mm chip won't be a straight linear decrease to 180 million chips. It may be only 100 million usable chips of that size. It will be a bigger decrease a 400 sq mm chip.

22july2013 said:

A tapered design where the monitor is basically thin at the top and thicker at the bottom could avoid the need for either a chin at the front or a bump at the back. And Apple has done it before with iMacs, which had a really fat middle, tapered to the edges.

No, A taper is unnecessary. The have a M1 processor in a 6.5 mm thick device that's been shipping for over 18 months now, and that device has a miniLED.

The only reason the chin is there is because they want it there. It's not a technical issue. They certainly can put a fan behind the display in a 11 mm thick device like the iMac 24 if they wanted to. They can put an M1 Pro or M1 Max in the iMac 24 if they wanted to.