In theory, with enough hardware and documentation so you don't confuse future you, you could put two different Splunk installations on a single server, and each installation controls a different cluster... But a single Splunk installation cannot be the cluster master for multiple different clusters. (It can control a single multi-site cluster, but this is rather different than having two separate clusters)
Simply structurally it cannot happen within a single installation first, server.conf is where ultimately you would put your cluster configuration, and there's a single [clustering] stanza for the cluster master to set its details (SF/RF) for the cluster. See http://docs.splunk.com/Documentation/Splunk/6.5.2/Indexer/Configuremasterwithserverconf and http://docs.splunk.com/Documentation/Splunk/6.5.2/Admin/Serverconf
Furthermore there's the problem of pushing configuration to the slaves which comes from a single $SPLUNK_HOME/etc/master-apps as well.
So technically you could have a single Node, it's not a single installation on a single Node.
