Hive supports 2 kinds of partitioning, injected and dynamic. You can totally use a partition key like the hour in UNIX time. It's an integer starting at some epoch and incrementing by 3600.

Now your query engine might require you to specify the partitions or range of partitions you want to query on; you absolutely can use datepartition >=a and datepartition<b in your query. Iceberg seems to fix that and just let you use the timestamp; presumably the metadata is smart enough to exclude the partitions you don't care about.