What are the advantages of serverless applications from Amazon?

Question

Amazon Web Services

I learned about serverless applications from Amazon, but the pricing principle is not clear.
There is a Compute Overhead and a Storage Overhead, and the compute is essentially running on its own regular server, with the IDE going to sleep after 30 minutes.
I thought it would be an IDE in the cloud with support for computing on powerful servers with 1Tb+ of RAM. And Ec2 instanse supports a maximum of 768 GB of RAM.
It's essentially a fee for an online collaborative development environment, right?
Moreover, the price is specially confusingly written so that people, without understanding, then pay forever, because the card is already linked.
And then the question arises, where can I cheaply try to make Python script calculations on clusters with 1Tb + RAM?

Answer 1

There is no serverless pricing confusion - it's a standard pay-as-u-go payment system. You actually pay for what you use. You don't need that much memory to run applications. Or rather, if you need it, then you have some kind of fig application, well, or such a specific case that you are in dedicated servers. Serverless is designed as a simple, integrable tool with a clear cost per number of hits and per processing time, rounded up every 100ms. What servers are running there? Believe me, they are powerful, but they won't tell us about it. The main bonus of being serverless is that it automatically scales without limits. That is, there may be several billion running copies of your code at the same time.
Nobody will give you cheap 1Tb RAM at all. You can make your cluster in AWS and run it only at the time of use. You don't pay when instances are turned off