‘live’ Integration of Pleiades Rack Saves 2 Million Hours
The NAS Division's recent "live" installation of a new rack on the Pleiades supercomputer has returned 2 million hours of computing time to users.
The new 512-core rack arrived in late December and installation was completed in early January. Integration into the Pleiades system was accomplished by connecting the new rack's InfiniBand (IB) dual port fabric via 44 fibre cables while Pleiades was running a full production workload.
This live integration saved 2 million hours in productivity that have previously been lost each time a planned system outage occurs. How?
Usually, when outages on Pleiades are planned, users get a one-week notice, and system utilization plummets about three days before the actual shutdown. This drop in usage is partly due to the fact that batch jobs are only started if they can finish by the start of the planned outage. About half of Pleiades' computational hours are consumed by long-running jobs; most take five days to complete further adding to the usage slowdown.
Looking ahead, use of the live integration process on additional Pleiades racks arriving in the spring will improve system availability for all users adding up to 8 million hours for mission projects over the life of the system.
New Cores do Twice the Work
The new rack contains Intel's Xeon X5570 (Nehalem) quad-core processors. Each core does 2.4 times the previous SGI Altix ICE generation's work and use of virtual cores in the Nehalem processors (16 virtual cores) yields even better performance returns for NASA's scientific and engineering users.
The January installation brings the total recent Pleiades expansion to 19 racks in October; 18 racks of SGI Altix ICE integrated blades were added. Pleiades now contains a total of 14,208 Intel Xeon 5500 series processors (56,832 cores in 111 racks).
Pleiades currently holds the 6th spot in the U.S. on the Top500 list of supercomputerswith a 544.3 teraflop/s LINPACK rating.
