[apothegm]

In addition to the network effects mentioned elsewhere: capital-intensive businesses that have raised significant capital and have a head start have moats.

So for instance, the reason a power company is a natural monopoly is that power transmission infrastructure is massively expensive to build, and the first one to raise the capital and build it out has a moat.

The same applies to something like video hosting that requires massive infrastructure (tho less so now than 15-20 years ago when YouTube was getting established.)

Or AWS — it takes massive capital to compete with them head to head, and even Google hasn’t entirely succeeded at that. (Microsoft sort of has, but mostly because their proprietary OS that’s so heavily relied on by enterprise is its own moat.)

Marketplaces are network effect adjacent. The reason no one really competes with Amazon (or in their niches, eBay or Etsy) is that all the buyers and all the sellers are already there. Same for Uber and Lyft.

IP is a type of moat for content based businesses like TV streaming services (Disney, Netflix, etc).

Apple tries to use the ubiquity of its devices as a moat for its services and massively profitable app store — though it’s a weak moat for the services and starting to be weakened by antitrust laws for apps/IAPs.

Technical moats, well… some tech expertise isn’t sufficiently widespread that any upstart competitor can hire someone to do the work. If you want to compete with Nvidia, good luck hiring all the world-class experts you’d need.

Microsoft’s moat? Switching costs, including learning curves. And all the software written specifically for Windows that’s not available on other platforms.

Switching costs are also a major moat for an email service like Gmail.

[al_borland]

> Apple tries to use the ubiquity of its devices

I think the ecosystem is Apple’s moat. The ubiquity of everyone in a friend group having them in since and its own moat that creates social pressure to have what your friends have, but I think the same could be said for Android.

Even without other people, if a person has more than one device, Apple’s integration between them is pretty solid. If someone already has one or two Apple products, it makes sense to stick with them for anything else, because it will all work together so much better. I got my first Mac because I had an iPod and the software for Windows was trash, I wanted iTunes and the idea of it being on Windows was just a rumor at that time. Once I had the Mac it made sense to get the iPhone. Once I had those it made sense to get the Apple TV, AirPods, Apple Watch… if I’m in the market for something in those product categories.

[deafpolygon]

Not to mention that Apple's approach (whether genuine or not) to consumer privacy, usability, etc. is perceived as better than Google's or Microsoft's. Branding is a moat in itself.

[moatel]

Nvidia mid or long term has no moat. It has to keep innovating. Groq are coming for them. AWS are coming for them. TSMC are in a country that could be at war in the next 10 years.

Guess thats a spicy take :)

[bobdvb]

Nvidia's Moat is CUDA, it's become so entrenched in development stacks that people have inertia in moving to anything else. I've been in situations where I've wanted to use other GPUs and everyone's said to be "but surely you want Nvidia?!" because it's the de facto now. AMD, Intel and others haven't done enough to break the NVidia API monopoly in the past decade. Intel's done stuff with One and AMD sponsored that CUDA converter thing. But there's not been enough effort to break through even when that's fundamentally the blocker.

If AMD and Intel want to compete, they need to think about the early years of computer software development. When people would say "Oh, we're not going to port our software for Amiga, PCs are now more popular." There is some portability in games for new GPUs, but it's CUDA in scientific and enterprise where they should have broken the CUDA stranglehold.