Getting Started on Edison
Before you can use or access Edison, you must have an active NERSC account and valid password. If you don't, see Accounts and Allocations.
Edison should be used in the following way: Log in to a "login" node, bring your files and data over, compile your code, and create a batch submission script. Then submit that script so that your application runs on the Edison "compute" nodes. Pay attention to the various file systems available and the choices in programming environments, and make sure you use the compiler "wrappers."
Edison is currently in its pre-production phase, a time when NERSC users are given access to the machine, and also when NERSC and Cray staff are still making improvements to the system. You are invited to run jobs on the Edison Phase II system free of charge in exchange for providing feedback to NERSC staff about your experiences using Edison.
The PGI compilers are not available on Edison. The default programming environment on Edison uses the Intel compilers. The Cray and GNU compilers are also available.
During the pre-production period the system may be taken down with little or no advance warning and support may be limited. During this period please:
- Report all bugs, problems, and usability issues you have with the system to email@example.com.
- Do not publish any machine performance results without the consent of NERSC. You may publish science results.
- Understand that your job may be aborted at any time if Cray or NERSC staff need to do testing or maintenance on the system.
- Back up your data often. The file systems are not yet being backed up.
Edison and Hopper have similar programming environments but there are some key architectural differences between them. This page describes those differences and how they may improve your productivity. Read More »