Edison
Status: Up
Edison is NERSC's newest supercomputer, a Cray XC30, with a peak performance of 2.57 petaflops/sec, 133,824 compute cores, 357 terabytes of memory, and 7.56 petabytes of disk.
Programming
Find information on how to compile your application, use MPI and OpenMP optimally, and use Cray's scientific libraries. Read More »
Running jobs
Learn how to submit a job to the queues and use the 'aprun' job launch command, monitor your job in the queue and job exit status. Also find information on OpenMP and MPI runtime settings as well as general application tuning options. Read More »
File Storage and I/O
The Edison system has 4 different file systems which provide different levels of disk storage, I/O performance and file permanence. Find information on I/O optimization, disk quotas and purging. Read More »
Debugging and Profiling
Learn about performance analysis tools CrayPat and IPM, debugging tools DDT and Totalview, and more. Read More »
Performance and Optimization
Learn useful optimization tips and view performance and benchmarking results on Edison. Read More »
Shared and Dynamic Libraries
By default the Edison system runs a limited OS kernel which does not support shared and dynamic libraries. However, users can enable the shared root file system by setting an environment variable (CRAY_ROOTFS=DSL) to enable shared and dynamic library applications to run on the Edison compute nodes. Cray-supported libraries such as SciLib, Portals, MPI, pthreads, and many others are available as dynamic shared objects (DSOs), and you can define and use your own DSOs, too. Shell and python scripts can also run on the compute nodes. Read More »
Cluster Compatibility Mode
Cluster Compatibility Mode (CCM) allows users to run standard ISV applications and other applications that need to use sockets on the Edison system. Read More »
Cray XC30 User Documentation
Cray documents describing the XC30 programming environment. Read More »
