I'm sitting at the National Science Foundation's quarterly review of Ranger which is now in
(hint, click on the "Resources" tab). I walked through the amazingly compact data center again early this morning. If a supercomputer can be called beautiful, Ranger is certainly that. What is probably most amazing are the
that have already been achieved in the first few months of use which several users have presented. But the numbers themselves are also amazing. Ranger has already amassed over 135 TB of data on its
file systems. The largest Lustre filesystem has a capacity of over 800 TB of user accessible space, can store over 470 million individual files (470M inodes), and file system metadata alone takes up 2 TB. More later, but turning back to the next speaker now.
ps, you probably want to know if Ranger passed the full system HPL (Linpack) benchmark of the
Top500 list. I'm not going to tell you what the exact results were, you will have to ask TACC, but I am happy to say Ranger passed the HPL target specified in the NSF acceptance tests.