ScyllaDB on AWS, One Large Node or More Smaller Ones?

[Guy]

I’m looking into running a ScyllaDB cluster on AWS. How do I determine sizing?
Is it better to run a few large instances or many smaller ones?