NERSC logo National Energy Research Scientific Computing Center
  A DOE Office of Science User Facility
  at Lawrence Berkeley National Laboratory
 

Franklin Quick Start Guide for Jacquard Users

This page provides some brief information for Jacquard users who are starting to use Franklin. Please see the other Franklin pages for more complete information.

CharacteristicsJacquardFranklin
Processor type 2.2 GHz Opteron 2.6 GHz Opteron
Processor/core theoretical peak 4.4 GFlop/sec 5.2 GFlop/sec
System theoretical peak (computational nodes) 3.13 TFlop/sec 101.5 TFlop/sec
Number of compute nodes 356 9,660
Processors/cores per node 2 2
Number of compute processors/cores 712 19,320
Physical memory per node 6 GBytes 4 GBytes
Usable memory per node 5 GBytes 3.75 GBytes
Number of spare compute nodes
(In service when possible.)
4 20
Number of login nodes 4 16
Node interconnect InfiniBand SeaStar2
File System GPFS Lustre
Usable disk space 30 TBytes 350 TBytes
Batch System PBS Torque/Moab

SSH Access to Franklin

SSH access to Franklin is the same as Jacquard. Franklin uses a particular ssh authentication method called keyboard interactive. End-user ssh software must supports this authentication method. Other common authentication methods are public key and password.

Passwords and Shells

Franklin, like Jacquard, uses the NIM password; it is also used for Bassi, PDSF, Davinci and NIM. Passwords cannot be changed directly on Franklin. NIM passwords are changed using the NIM web account interface. NIM is also used to change the default login shell on Franklin.

Charging

Charging is in units called MPP hours. One Franklin hour is equal to 6.5 MPP hours. See Accounts and Charging on Franklin.

Compiling and Linking

The default compilers on Franklin are Portland Group Compilers (PGI). GNU compilers are also available. The default compilers on Jacquard are Pathscale compilers. This table shows the comparison for most commonly used Pathscale and XT4 Fortran compiler options.

The Cray provided compiler wrappers (ftn, cc and CC) should be used for serial and parallel jobs on Franklin which will be run on compute nodes. The native compilers (such as pgf90, pgcc, gcc, mpif90, mpicc, etc) are not recommended due to the complexity of the environment.

The default environment on Franklin uses 64-bit addressing, just like Jacquard.

Please see Programming on Franklin for more information.

Running Jobs

Franklin uses Torque/Moab for batch job schedulers. Jacquard uses PBS Pro/Moab. Batch jobs should run in user's $SCRATCH directory (instead of $HOME) for better I/O performance on Franklin. Please see Running jobs on Franklin for more information.

File Systems

Franklin users have a $HOME directory and a $SCRATCH directory. The /projects center-wide file system is also mounted on Franklin. Both $HOME and $SCRATCH on Franklin are Lustre file systems, while on Jacquard they are provided by GPFS file systems.

As on Jacquard, the $HOME directory and default file group are the personal file group which has the same name as user login name.

Reading Binary Data from Jacquard

Both Jacquard and Franklin are "little endian" machines. Being little endian means the lower-order byte of a number is stored at the lower address in memory while the higher-order byte stored in higher address. It is reversed for a big endian machine. Binary data from Jacquard should be readable on Franklin.


LBNL Home
Page last modified: Fri, 09 Nov 2007 21:19:21 GMT
Page URL: http://www.nersc.gov/nusers/systems/franklin/for_jacquard_users.php
Web contact: webmaster@nersc.gov
Computing questions: consult@nersc.gov

Privacy and Security Notice
DOE Office of Science