Apple Vision Pro Musings
engineered.ooo> I’m not saying they couldn’t produce something like Vision Pro if they wanted to—it’s just not their priority.
Nothing that Facebook has ever done gives the slightest indication that they're capable of creating anything like the Vision Pro.
Just for starters, where do they get a processor close enough to the M2's performance per watt? Where do they get an operating system to drive it? Who creates a chip like the R1 that handles the video from multiple cameras and gets it displayed in 12 milliseconds?
So much of the technology had to be created from scratch, based on Apple stating they filed over 5,000 patents related to the Vision Pro.
The list of companies that could have pulled this off if they wanted to and had the resources to invest is pretty short--there's no way Facebook is on it.
I'ma huge fan of Apple and the Vision Pro as well as Apple Silicon. But it's a bit short-sighted to say that Meta or other companies can't make the same thing.
First of all, Qualcomm's SoCs, while slower, can and will approach M2's performance in about a year or two - especially if they can get Nuvia's designs out the door.
Second, we don't know about the R1 and how impressive it is. But I doubt it's some kind of magical chip that only Apple can make.
Third, the patents won't really do that much because many tech companies have similar/overlapping patents and Apple and Meta won't go to war over patents. The only winner there would be lawyers.
Fourth, I'm going to bet that it won't be Meta who will successfully copy Vision Pro. It'll be one or more of the Chinese companies. Some of them have already created some impressive true AR glasses - not headsets.
> First of all, Qualcomm's SoCs, while slower, can and will approach M2's performance in about a year or two - especially if they can get Nuvia's designs out the door.
The M2 is using TSMC's 5nm process; we know the M3 is already in the works using 3nm, which means less power and more performance than the M2. Even if Qualcomm succeeds and is able to create a processor as fast and power efficient as the M2 is now, it won't be as fast or as power efficient as the M3.
> Second, we don't know about the R1 and how impressive it is. But I doubt it's some kind of magical chip that only Apple can make.
Nobody said anything about it being magic or that only Apple can create such a chip. But the list is short…
The R1 handles the processing of the video from the Vision Pro's 12 cameras and 5 sensors. Two of the cameras transmit more than a billion pixels per second to the display which has a refresh rate of 90Hz, with a peak at 96Hz. The R1 is able to stream images to the display within 12 milliseconds [1].
> Third, the patents won't really do that much because many tech companies have similar/overlapping patents and Apple and Meta won't go to war over patents.
Not sure if you've been paying attention but Apple is constantly in patent disputes with other companies. There's no way after investing an estimated $15 billion in R&D costs would Apple allow Meta or anyone else infringe on their Vision Pro patents.
> Fourth, I'm going to bet that it won't be Meta who will successfully copy Vision Pro. It'll be one or more of the Chinese companies.
I'm sure there will be plenty of bootleg Apple Vision Pro headsets for sale in Asia fairly soon… they will look like the Vision Pro but they won't function like the Vision Pro. And they certainly won't have the operating system or the ecosystem that Apple has.
[1]: https://www.macrumors.com/2023/06/07/apple-vision-hardware/
1. Qualcomm will also have access to TSMC’s 3nm tech. Given that the Vision Pro will launch with M2 in 6 months, it might not get M3 for 1.5 year. That’s plenty of time for Qualcomm to catch up.
2. Again, I’m not convinced that the R1 is special. I think other companies will easily duplicate it.
3. Apple doesn’t usually go around suing other huge large tech companies. They won’t sue Meta because they have a ton of AR and VR patents too.
4. Chinese companies make some insane hardware. I think they’re much better than Meta at making hardware. And they’ve already demonstrated some very advanced AR devices.
> Qualcomm will also have access to TSMC’s 3nm tech. Given that the Vision Pro will launch with M2 in 6 months, it might not get M3 for 1.5 year. That’s plenty of time for Qualcomm to catch up.
I wouldn’t assume Vision Pro will ship with an M2… it could be a M2 Pro or another variation.
If past is prologue, Qualcomm has yet to match Apple Silicon performance. Sure Qualcomm will have access to TSMC’s 3nm tech but they don’t have the ability to pay billions of dollars in advance to reserve essentially all of TSMC’s 3nm capacity [1] as Apple is rumored to have done. So even if there was the demand, Qualcomm won’t be able to make enough chips.
Probably more so than most companies, Apple plays the long game.
Finally, when Steve Jobs stated Apple had filed something like 300 patents and a 5-year lead on every other company, the industry scoffed. RIM’s founders doubted the specs and Steve Balmer laughed at the thought of the iPhone competing with Microsoft, never mind Nokia, RIM and Motorola. I don’t have to mention what happened to these leaders of the smartphone industry.
For what Apple is attempting to accomplish, they’re essentially as far ahead in 2023 as they were in 2007. I know that sounds crazy but that’s how Apple rolls.
Even if a company could create comparable hardware, they wouldn’t come close regarding software and integration into an existing ecosystem of iOS + macOS software, not to mention VisionOS apps.
[1]: https://www.macrumors.com/2023/05/15/apple-tsmc-3nm-producti...
>*I wouldn’t assume Vision Pro will ship with an M2… it could be a M2 Pro or another variation.
It has 2 hours of battery life using the M2. The M2 Pro has 2x the power requirements at full load. There's no point in putting an M2 Pro in it.