Nothing will happen if you hibarnate, you can click continue and that’s it.
1600 items in 10 hours? There is something very wrong with the database there. I just tested a single core virtual machine with 512MB of ram with 50 000 items, and it indexed in like 45 minutes and almost 0 CPU load, that should be a piece of cake for that machine.
The server specification is definitely all right, that is a super high specification for such small amount of data.
Do you have any strong data replication or data restoration service running on that server? I mean a service which lets you roll back to any point in time on your database?
That could explain the problem, as it has to store a a lot of redundant data before every single database row insertion. The plugin tries to insert thousands of rows and the server might be trying to remember and log all of the changes and it basically kills the CPU.