another syntax question

[Eric Miller]

For non-parallel applications, is it possible to run individual instances on
diskless nodes?  For example, I want to execute a non-MPI program "A" that
is located in the /bin directory of my master node, but I want to run one
instance of "A" on each of my diskless nodes.

What is the syntax that equates to:

#NP=1 "A" on node0 only
#NP=1 "A" on node1 only
#....
#....