Hmm. A bit more information is necessary here. What's the size of tempdb when it's grown out quite a bit? From your settings, it looks like tempdb is configured to grow in very small increments; typically (though this is a very "it depends!" sort of thing), tempdb will be decently large, depending on your environment.

This is largely because tempdb is the crux of many operations that get performed in queries, such as sorts, temp table allocation, and tons of other things. Those operations, if called enough, will cause growth, sometimes in very large amounts. I believe even the internal workings of SQL Server will be using tempdb quite a bit.

For further investigation, I'd recommend reading SQLServerCentral's Stairway to Server-Side Tracing:

Stairway[/url]

I'd suggest setting up a server-side trace and running it to capture what's running on the server, particularly when you say there's no queries running; it may be internal operations causing tempdb strain, or there could be other unknowns that could be influencing tempdb usage.

Let us know what you find, and someone will probably be able to offer assistance!

(On a side note, interesting user name! Play Marvel 2 often? :-D)

EDIT: Well, nevermind! A trace might be a bit overboard (though still intriguing); did some digging, since the topic sparked my interest. Here's a great article by Jeremiah Peschka about figuring things out about tempdb:

Article[/url]

In particular, the last query in the article should match up to the results from the query you've already done; from there, you can take the contents of the sql_handle column and plug them into sys.dm_exec_sql_text() between the parenthesis to get the exact query that caused the usage.