Another day, another Twitter awesome moment, Parikshit Savjani was explaining how backups are now faster in SQL Server 2017.

For more info on how, read his post. TD;LR - The wait time before data is actually copied is reduced because the amount of buffers to scan is massively reduced.

“…requires only 250 buffers to scan as opposed to 500 Million buffers with former algorithm…”

All newly created database on SQL 2017 will get this benefit, existing will require a little work

alter database set target_recovery_time = 60 seconds