Because you are pooling computer resources across many more users you have better amortized cost than if you had a workstation on every home that was idle 99% of the time. Of course, latency and bandwidth to the cell tower is worse than wifi, but better than if the compute is done on a remote server.
I have no idea of which model will succeed for which use cases, but the idea is sound.
Again: which computation paradigm works best depends on the use case. It is not an all or nothing situation.
As much as there are market drivers to make the cloud attractive to businesses and legitimate reasons why cloud solutions are better than alternatives there is a real desire in a growing number of people to have a solution that they can tinker with but that also has a polished UI so they don't have to tinker with it when they don't want to.
I think there are many early innovators that fail in later stage growth because of this issue.
I'm still surprised they did not create an App Store for AI. Basically lock everything down and make developers pay a % of their revenue, Apple style.
At enterprise scale, locked down marketplaces don't work. They act as an forcing factor for larger organizations to build in house because no one wants vendor lock-in or to lose money via an arbitrage.
This is a major reason why you'll see large deals pushing for enhanced customization options or API parity, as larger customers have the ability to push back against vendor lock-in.
Furthermore, a relatively open market (eg. NGC) acts as a loss-leader by allowing a community to develop using a corporate standard, thus allowing you to build stickiness without directly impacting a customer's bottom-line
Fundamentally, a company driven by Enterprise revenue (eg. Nvidia) will have a different marketplace structure from a B2C product such as Apple's App Store where purchasers have little power.
More tactically, excessive charging on marketplace pushes vendors away from selling on AWS Marketplace and makes them develop alternative deployment methods, which reduce the stickiness of AWS, as hyperscalers are commodified nowadays.
Motorola learned that the hard way 40 years ago when pushing excessively restrictive OEM and Partnership rules compared to IBM.
AWS is only as strong as it's Partnership ecosystem, as companies that are purchasing tend to use 80-90 different apps along with their cloud.
Basically, Enteprise Sales shows hallmarks of a Stag Hunt Game, so a mutually beneficial pricing strategy amongst vendors (AWS, AWS Partners such as Nvidia, MSP) is ideal.
If this was true, we'd have more mobile computing platforms. Large enterprises publish in Apple's AppStore.
Purchasers from Apple's App Store are primarily individual consumers. It is a B2C play.
Monetizing an Nvidia marketplace such as NGC would be foolhardy as the primary users/"purchasers" are organizations with budgets and procurement power. It is an Enterprise B2B player.
In enterprise sales, the power differential between (mid- and upper-market) customers and vendors is in the customer's favor, as they have significant buying power and thus a higher user acquisition cost. The upside is revenue is much higher, margins are better, and you can differentiate on product as commodification is difficult.
This is less so in consumer facing sales as customers have significantly weaker buying power, but conversely have a much lower user acquisition cost at scale. Hence, a growth-based GTM approach is critical, as you need customers in aggregate to truly unlock revenue at scale.
Also, I don't believe the balance of power is tipping towards business customers, as nvidia is basically the only relevant player.
If you think smartphone customers and server customers are evaluating hardware based on the same criteria, then why isn't Apple the leading datacenter hardware OEM?
> Large enterprises publish in Apple's AppStore.
Yeah? Where's my Pro Tools download on the App Store? Where's my Cinema4D download? Can I get Bitwig Studio from there? Hell, is iTerm2 or Hammerspoon even available there?
Large enterprises very explicitly don't publish on the MacOS App Store because it is a purely raw deal. If you're developing a cross-platform app (which most large enterprises do), then you've already solved all the problems the App Store offers to help with. It's a burdonsome tax for anyone that's not a helpless indie, and even the indies lack the negotiating power that makes the App Store profitable for certain enterprises.
Because Apple doesn't want to be in the B2B space.
But you can say exactly the same thing about large companies publishing (consumer apps) in the App Store. Why would they want vendor lock-in?
Their bet is that AI will unlock robotics use and they don't want to be simply compute providers, they want to innovate on the whole chain, software, hardware, services, everything.
Their position is quite unique as their R&D is basically financed by their future competitors, they are making bank while going where the puck will be.