This looks at how cloud computing evolves, typical use cases and players in the "cloud wars". ## Consumer Cloud The personal cloud is essentially a mainframe -> terminal relationship, meaning that the user offloads computing power to a centralized system that provides additional capabilities to the client machine. However as we see with [[Edge Computing]] the devices accessing the cloud get progressively more powerful and capable, thus not requiring the cloud computing power of capabilities anymore. Example: "The cloud is able to give you GB of data storage!" vs. "I now have TB of data storage locally." Alternatively: "The cloud is able to give you any graphics card you want!" vs. "My local graphics card is able to run pretty much any AAA game anyway." This is including the technical and UX challenges the cloud inherently has when trying to solve local consumer concerns, for example cloud storage requiring massive network data and bandwidth caps or cloud gaming requiring insanely low latency to be anywhere near viable. So for most personal / consumer cloud systems there is a natural limit / time box of usefulness until the consumer grade devices match or outpace the cloud components. This is also the reason why I think cloud gaming does not make sense as a way to save costs for expensive gaming hardware. There are exceptions where for example the cloud would render / compute things that are outside the realm of consumer class hardware, for example Flight Simulator 2020 simulating a global weather model in the cloud, streaming the data to PCs. That said, there are enterprise cloud use cases that are also interesting for consumers, for example off-site backups and central data management. The same is true for the [[#Cloud as Application Server SaaS]], for example Office online. Unsurprisingly those consumer cloud cases are owned by companies creating consumer products, which then tie into their cloud services, usually for an additional fee. For example Apple, Google and Microsoft with "lifestyle" services like photo and data backups (Google Photos, Apple iCloud, Microsoft OneDrive) or gaming (XBOX Game Pass, Apple Arcade, Google Stadia). ## Enterprise Cloud The enterprise cloud is essentially about outsourcing your local data center hardware and management to somebody else. Before organizations would have their own servers running *somewhere* (aka under the desk of "that admin person"), now they offload all server tasks to an enterprise provider. The cloud was a pretty natural progression for enterprise as they had local data centers in every location -> central data centers in some locations that all other connected to -> central data centers managed by others, usually their web / online providers -> all data centers managed by the people that do that online stuff. From there, enterprise cloud capabilities evolved to allow the cloud providers to manage their compute power more effectively. Sure, serverless compute is great for developers, but it also allows to slice and distribute customer jobs more effectively. Essentially it's similar to defragmentation of hard drives, only with computing power. Within enterprise clouds there are two major lines of thinking: * "I want to use the cloud as I would my own machines and do my own DevOps" -> A lot of manual control, high maintenance, high skill -> [[Amazon]] AWS rules here * "I want the cloud provider to take care of the DevOps while I'll focus on development" -> Limited manual control, medium maintenance. low skill -> Azure rules here [[Google]] cloud is somewhere in between I guess? It's mostly on the "do your own DevOps" side, but cheaper than AWS and it does cool things with VMs. Also the inter-cloud network is pretty fast.