testpara1_lapw can be used to check the progress of lapw1. I typed this command and got:<br>ec5-21: not running<br>ec5-21: not running<br>ec5-21: not running<br>ec5-21: not running<br>ec5-34: not running<br>ec5-34: not running<br>
ec5-34: not running<br>ec5-34: running<br>(this is only part of what I got, there are another 72 cores not running)<br>However, when I ssh to ec5-34, there isn't any job running. I then ssh to ec5-21, I found lapw1 running. Is there a reasonable explanation for this? Does wien2k scale well with a GigE network? It seems I have to wait for one core to finish while the other 79 cores are idle.<br>
<br><br>Yundi<br>