Inefficient Write+Copy+Delete pattern when writing to S3.

[apghml]

When tf.io writes a file, it checks whether HasAtomicMove is true and if so, simulates atomic writes by first writing to a temporary file, then renaming the file to the correct name. This is great on a local filesystem. But for S3, this is undesirable behavior for a few reasons:

1. S3 writes are already atomic, so there is no need to simulate one.
2. In the AWS S3 SDK, which tf.io uses, moving a file is implemented as a copy+delete, which increases the load on S3 compared to a direct write.

[learning-to-play]

Hi @yongtang , This issue has is preventing some users from running their jobs on GPU by causing additional load. Could you help?