Overview
The ETH High-Performance-Cluster (HPC) Euler is a powerful cluster and it is available to all ETH researchers and students.
Leonhard Euler (1707-1783) is a Swiss mathematician, physicist, astronomer, logician and engineer who made important and influential discoveries in many branches of mathematics.
The cluster is based on a shareholder model and is therefore financed by its users. There is a share financed by the IT Services available to all ETH researchers and students at no cost (public share). Since early 2016 the GDC is shareholder of Euler. GDC users can become member of this share and thereby benefit from its resources.
Advantages of using HPC Euler:
- The cluster (hard- and software) is professionally managed by the Informatik-Dienste
- Computing peaks can be accommodated
- Storage needs can be accommodated flexibly
Advantages of being part of the GDC-Euler share:
- Access to maintained GDC software stack
- Higher priority to submit jobs
- Support from GDC bioinformaticians
Our GDC share does not contain any GPUs.
- Login to a login-node via your computer.
- Submit your jobs on the Login nodes. Jobs get sent to the Compute nodes by the scheduler.
- The jobs run on the best compute node that becomes available and results will be saved in your your working directory (Scratch).
- Collect the results (select, filter, compress etc.).
- Transfer final results back to your computer.
Rules
We need rules and commands to guarantee fair access to the share and to make sure all the jobs submitted daily can run smoothly on the cluster. Therefore, it is important that all users understand the basics, comply with the rules and follow instructions given by the GDC bioinformaticians. We intend this manual to provide guidance for GDC users that would like to use the GDC share on Euler. Please read the manual carefully and ask for help if needed.
Contacts
Although the HPC Euler is managed by the cluster support, the GDC bioinformaticians are your first contact regarding Euler usage, problems and software requests. We do our best to find solutions and answers for your problems as quickly as possible. If necessary, we will involve the cluster support team.