Hacker News Reader

Comment by Retric | Hacker News Reader

The Vision Pro’s existing sales have already justified the project. They have sold 560+ million dollars worth of them at presumably insane profit margins.

I think Apple’s strategy is to just execute whatever the next tech thing is better than existing participants. MP3 players weren’t a forever thing, but they were a very profitable thing for Apple. Watch, tablet, 3D glasses etc timed after someone else proved the market and just in time for Apple to slurp up the most profitable slice of the pie.

They don’t want to be the first self driving car company, they want to be the most profitable self driving car company.

View on HN

Apple's profit margin on the Vision Pro, from a purely bill of material perspective, is estimated to be around 45%. But of course they have spent billions of R&D towards that project and it is far from out of the woods yet. It will be in the red for a few iterations over years.

Meta has sold more than twenty million Quest headsets of various types, and by most measures that project is still basically a money pit.

That’s 20 million headsets with ~zero to negative profit margin on each sale.

Meta’s investment in VR has been a dumpster fire, but the headset R&D isn’t where the overwhelming majority of that money is going. They’ve done everything from making VR movies to promoting 3rd party games most recently spending 400 million for beat saber’s publisher.

They bought beat studios back in 2019.

More recently they also bought Supernatural (fitness app)

Meta are playing the long game with this, they are building out the eco system and the devices. They are doing a great job despite it being a money pit, the Quest 3 is an awesome device. It just needs a bit better software and better marketing around those killer apps, like fitness etc.

I don’t believe for one second that they are doing a great job. They have lost so much money. The only reason they haven’t completely abandoned the idea is because Zuckerberg has a controlling stake in the stock.

Meta is investing in VR movies? How much and where? I can count on my fingers the good VR videos from Meta, and I mean that from a tech point of view. The median VR porn has better video quality than 90% of videos on the Meta video app. Also Youtube VR is limited to just 4k and looks like garbage. VR video only starts to look good from 6-8k and 60fps.

Afaik they funded a substantial amount of the Felix and Paul films.

They do fund VR movies. I worked in 2015-17 in the area and they were doing it then.

Meta quite likely also subsidize their headset. They follow the console market plan of getting money back on software sales, but a big issue has been incredibly low user retention.

With low retention, every sold device is likely costing them.

I'm not saying Vision Pro won't eventually justify itself, but $560m revenue is likely nowhere near the cost for this programme so far.

This project has been rumoured to have hundreds of engineers for 5+ years, that's that revenue already (let alone profit), then the manufacturing setup, new chip development, etc.

Even if this product makes $2bn revenue, which could be unlikely just based on the rumoured manufacturing limitations and number they expect to manufacture this year, it probably wouldn't be justified once you figure out the profit margins. This is clearly all about the next product, or the one after. It's about the iPhone 4 launch, for this product, not the first, 3G, or 3GS practice runs.

Apple have a history of healthy margins, but huge production costs on their high end stuff, so while this is $3500, I wouldn't be surprised if it's still only 50% margin. That's great, but it's not the 70% one might expect from such a high price, and it will be because of things like maybe the body costs $200 rather than $50 because of some silly uncompromising design choices that don't make sense for high volume consumer hardware.

I doubt they’re into the billions on R&D.

The project could be 7-10 years old but they didn’t start on day one with hundreds of people. Let’s say a core of ~50 people for 5 years as proof of concept then ~350 people for another 5 years at 300k per person on average, that’s a 600 million dollar investment.

Sure they could have a larger team, but they also don’t pay testers anything close to 150k year.

Now discuss the millions spent on the physical R&D, corp acquisitions (Apple was for a while averaging 1-2 acquisitions/day), whatever licensing fees they need to pay out for initial access, this is easily a multibillion dollar bet.

Try and put some actual numbers to those supposed investments such as Vision Pro specific acquisitions…

$300k/annum is the carrying cost of low to mid engineers at big tech. For the kind of talent you need to pull this off, I’d at least double it (and maybe closer to $1M/year for the early folks)

Employees are expensive. That $200k salary comes with 100k in RSUs and then 25+% overhead.

Then factor in bonuses for leadership when they finally get the thing to ship…

And 350 people sounds like way too few, considering the novel R&D. Maybe 2000 would be my guess.

What’s the ratio of testers:developers and how much do testers make? Not everyone on a this kind of a project is making even low end software developer salaries, and not every RSU vests which is a big part of why they are so popular.

I’ve heard people quote hundreds of people working on the Vision Pro which is already a lot of people for Apple. The company only recently broke 150k employees globally and a large fraction of that are working retail at their stores. Now split the remainder across iPhone, Mac, iPad, App Store, Apple TV, Air Pods, Watch, marketing, this car project, back end infrastructure etc and ~2,000 people on Vision Pro for 5+ years just doesn’t line up.

I agree that for the Vision Pro headcount 2000 is unlikely, but you also have to include all the other functions – chip design, marketing, all the people that interface with the manufacturing partners, the manufacturing partners themselves, the people who design the production lines, the people who design and program the machines on the production line, the people embedded in TSMC overseeing chip fab, the people TSMC will have embedded in Apple overseeing chip design... it goes on.

For an idea of the scope, a number of years ago it was reported that Apple had 40 business class seats from SFO to Shenzhen pre-booked every single day. That's how much it takes to interface with manufacturing. Now this isn't the iPhone, the scale is smaller, but this stuff is hard and takes a lot of people, it's practically why Tim Cook got the CEO job.

I completely agree manufacturing takes significant oversight, but only a tiny fraction of that manpower goes back 2+ years on a small scale production project like this.

Apple has a pipeline and they move experts between projects. The team making commercials is another late yet critical addition.

IME, having worked at big tech on moonshots, outsiders wildly underestimate how many people it takes to build a new platform. There is just so much work needed. I could be wrong on the AVP, Apple might make it work, but my prior is what it is.

Rumors (aka Bloomberg) have it that Apple spent about $1bil per year on the headset: https://www.yahoo.com/entertainment/apple-bid-fuel-vr-landsc...

Admittedly, some of the tech was used in other projects too, such as spatial audio or LIDAR, and I'm not sure where that's counted.

A lot of the individual tech pieces to make it work have been going into other products for years. Like head tracking in AirPods, face id, Memoji, probably lidar, certainly other things that aren't coming to me right now. A lot of things that seemed like gimmicks that turned out, it seems to me, were really for Vision Pro.

I highly doubt that the head tracking on AirPods is the same implementation as VP. VP uses iris ID, not Face ID. Memoji is a non factor. Lidar’s usage is probably also very different.

There’s minor overlap but I feel like the R&D cost of the VP didn’t have much overlap with other products. Except the M2 SoC.

Yea but the inherent costs are not in the implementation. The complexity of integrating something like head tracking is only 10% to do with the “type” of head tracking.

So even if VP head tracking works differently, the groundwork has been laid for that category of feature.

Why wouldn't the head tracking be the same? They use the same H2 chip as AirPods. FaceID probably implemented ideas used by iris ID. Learnings from Memoji almost definitely underly the personas. Lidar in the wild I am sure provided valuable data and got developers used to working with AR.

From what has been reported - a lot of the AR things, including RealityKit and ARKit came from the TDG which are also who made VP.

Meta burned through ~$50B to get where they are with Oculus. Apple has likely spend at least a reasonable fraction of that. My guess is somewhere in the league of single digit billions.

Developing R1 itself would have cost ~$500M. Not to mention the cost of tooling the factories for mass production. Apple has fronted money to TSMC in the past to accelerate their manufacturing roadmap.

That said - a car is a completely different beast. They are going to upend a lot of things with it, including the dealership model. Probably timing it based on projections of when the demand is set to rise, given many countries are “banning” sale of new ICE from 2035. Though my money is on that deadline getting extended a few times.

Tesla already upended the dealership model.

But the point is: Tesla had to build a service network. Apple would have to find a solution for that too. You cannot bring your car to an Apple store nor mail it in, if something is broken.

It’s unfathomable to me how $50B are spent.

I seriously doubt Oculus VR would have spent 500M as an independent company on the R1. Especially when looking at the other head set manufacturers.

Mega wasted 50B on VR as a category but that includes money subsidizing headsets, making VR movies, buying an VR company that had less than 100M invested for 2B etc. It’s a story of management incompetence not some requirement for headsets R&D.

Untold R&D billions went into this; we have no idea what the breakeven would be.

If we were to treat Apple and Meta equivalently, we would have spent a decade tallying Apple’s R&D investments as “losses”.

I think the phrase you're looking for is second-mover advantage:

https://en.wikipedia.org/wiki/First-mover_advantage#Second-m...

I am looking forward to seeing Apple execute on Generative AI. While I think shoving it into every product you can think of (Google, Samsung) might have unforeseen consequences, I don’t think the other extreme is the right play either. This field is moving fast and they may not have time to do it the Apple way.

If they execute the Apple car like they executed the Apple Watch, then the car will look like:

https://en.wikipedia.org/wiki/Fiat_Multipla

But yes, everybody will love it.

Who proved the market with the tablet before the iPad?

The iPad was initially just a bigger iPod. It wasn't that much of a leap.

> The iPad was initially just a bigger iPod. It wasn't that much of a leap.

wrong ancestry. iPad from Newton.

What specifically do you think the iPad inherited from the Newton, that didn't come to it through the iPhone/iPod Touch lineage?

> What specifically do you think the iPad inherited from the Newton, that didn't come to it through the iPhone/iPod Touch lineage

iPhone / iPod Touch / iPad all come from the Newton lineage.

But the person I responded to said “iPod”. The iPod is far removed from iPhone / iPhone Touch / and iPad in terms of interaction modality, operating system, application platform, etc.

The iPod doesn’t run iOS, is a media player not general purpose computing device, like the others and their ancestor, the Newton.

iPod touch came out in 2007 and ran iOS, 3 years before the iPad.

iPad was basically a big iPod touch, which was basically an iPhone with no modem.

The person you responded to did say "iPod" but they surely meant iPod Touch. It's an easy mistake to make, you even made a similar mistake above writing "iPhone" when you meant "iPod" so very easy to do:

> iPhone / iPhone Touch / and iPad

I have both of those devices on my desk. The originals of both. The similarities of both: rectangular, displays, batteries (dead), audio jack, charging ports and could play music.

Aside from the large list of physical differences (like, one has cameras while one has a big flat physical turntable as its input device), the biggest category of differences is as large as the number of unique apps there are in the app store. Because the biggest difference is: one does anything, the other does one thing.

Still, you're right. One doesn't leap over oceans of differences. There were many products and ideas between the original iPod and the iPad, each one potentially sharing with or influencing the other.

Pretty sure they meant a big iPod touch, which was basically an iPhone without the phone part, not a gen 1 iPod.

The 2nd gen iPad was the real magic, and made me start paying more attention to Apple. There were rumours of a "Retina display", higher res than an HDTV in a tablet size. I was doubtful. I figured if display manufacturers could do that, we wouldn't be stuck with all these 1080p monitors and 768p laptops... Well, it turns out I was wrong. Display manufacturers just can't be bothered to make compelling displays on their own for some reason.

That crisp display, with enough horsepower to run things smoothly, there was nothing else comparable at the time.

Not sure who downvoted but OP is right, if you understand it was the iPod Touch.

Arguably Palm, but Microsoft popularized the name tablets as a category in 2000 even having “Windows XP Tablet PC Edition” in 2002.

However, various products gained traction as the hardware was able to hit a better price point.

2005: https://en.wikipedia.org/wiki/Nokia_770_Internet_Tablet 2007: https://en.wikipedia.org/wiki/Nokia_N800

Where does the $560M number come from?

analyst estimates are that they sold about 160,000 units through preorders. At $3500 per unit, that's $560M.

> And the supercharger network is no longer an advantage as almost all EVs will now be using them.

Having more demand for the network means the network can grow and expand at a much faster rate and make it cheaper. That's good for Tesla drivers in the long term. And Tesla itself. Tesla making more $$ also filters back to the consumers through R&D and new products.

If EV is truly going to be everywhere then charging needs to be generic and everywhere like today's gas stations. Currently it's still very much a luxury type thing when you consider a lower-middle class person living in a skyscraper in a city (which is a ton of people these days). In urban areas you don't have the luxury of multiple separate gas station services, given the limited real estate. Having to drive longer distances to use one near your apartment for ex. The more the better.

The network was always going to become a commodity anyway. Why not establish yourself as the leader ahead of time?

Did they spend less than 500M to develop it?

Reply on news.ycombinator.com