Selecting one or more cloud vendors is a key choice when adopting public cloud services. I have done most of my production work with AWS or GCP. However, for some scenarios, Azure or Alibaba Cloud are 'best fit' choices as well.
The global market trends (shown below) can be seen as a starting point for you.
It is important to further refine your selection based on the following:
- services needed
- does a potential vendor offer your desired services in their cloud?
- comparing service x vendor, which fits best?
- example: does
AWS IoTorAzure IoTbetter meet your needs?
- services location(s)
- does a potential vendor have your desired services in your preferred location(s)?
- where are their data center(s)?
- community cloud skills
- what is the size of the total technical community for a potential vendor?
- what is the size of the local technical community in your geography for a potential vendor?
- Positive - easy to set up for dev/experiements, cheap/fast & scalable, innovative ML services
- Negative - Enterprise setup/support complex, idiosyncratic (i.e. 'Googly') service implementation
- Positive - market share/ partner ecosystm, huge variety of services
- Negative - dated, security/cost management is arduous
- Positive - Active Directory integration simplifies security, core services are solid performers
- Negative - expensive, must test scalability by service
- Positive - good variety of services, well-priced
- Negative - derivative of AWS, based in China (privacy concerns), requires US passport (or other ID) to setup account
- Positive - innovation in quantum computing - great tools, useful Quantum libraries and tools
- Negative - expensive for general cloud service