I haven't been crunching that long on this project, but I've noticed that several tasks complete in 1-2 seconds. These tasks have the same estimated computation size as the tasks that take 1.5-2 hours, but they still show as validated, even though they only get 0.01 credit or something like that.

What are these tasks doing that they run for 2 seconds? I'm not the only one; I've browsed the top users and top computers list, and a lot of people get these 2 second tasks mixed in.

I haven't been crunching that long on this project, but I've noticed that several tasks complete in 1-2 seconds. These tasks have the same estimated computation size as the tasks that take 1.5-2 hours, but they still show as validated, even though they only get 0.01 credit or something like that.

What are these tasks doing that they run for 2 seconds? I'm not the only one; I've browsed the top users and top computers list, and a lot of people get these 2 second tasks mixed in.

This has been discussed before, but it's buried somewhere in the old posts.

In a nutshell, the work generator breaks the search space into equal pieces. Sometimes the pieces are really quick. There's no easy way to determine how long a given WU will take without actually running it.

For the bounded app, where >50% of the WUs are fast, I run the WUs through a filter to weed out the really fast ones (<10 sec). At some point I may need to do this for the standard app.