ICTP cloud

A the participants of the computational school on "Ab-initio Many-body Methods and Simulations with the Yambo Code" will be sent a link and a password to connect via noVNC to a Virtual Machine running on the ICTP cloud server. This is the strongly suggested mode to work on the hand-on sessions. noVNC is a open source VNC (Virtual Network Computing, a graphical desktop-sharing system) client that runs well in any modern browser.

How to connect

• Click on the link that you received by the school organizers or copy and paste it in the url bar of your favorite browser.

• Click on the connect button and insert the password.

Your browser will become a linux desktop

How to load the Yambo code and the tutorials

(i) Open a terminal (click on the black terminal icon, the second from the left in the lower part of the screen) and load the yambo code via spack:

spack load yambo

(ii) The tutorial material is available for the copy from this path:

cp -r /media/ictpuser/smr3694/ictptutor/YAMBO_TUTORIALS ~/

(iii) One of the tutorial shows the usage of Yambopy, it is installed via Anaconda. Yambopy is used both to manage workflows and to do postprocessing. So, in order to use it you have to load three spack pakages:

spack load quantum-espresso
spack load yambo
spack load anaconda3

Other useful spack commands

In the virtual machine Yambo was installed via the Spack tool. Spack is a package manager for supercomputers, Linux, and macOS. It makes installing scientific software easy.

spack find [options] [package]

List and search installed packages.

Useful options can be:

-p: show paths to package install directories
-v: show variants in output (can be long)

An example:

ictptutor@smr3694test:~/YAMBO_TUTORIALS$ spack find -v yambo
==> 1 installed package
-- linux-ubuntu20.04-cascadelake / gcc@11.2.0 -------------------
yambo@5.1.0~cuda~dp+mpi+openmp~parallel_io+ph+rt~yambopy cuda_arch=none linalg=slepc 
patches=b9362020b0a29abec535afd7d782b8bb643678fe9215815ca8dc9e4941cb169f,e3dcb3df39e2c70ebd57ded322d4ddbc1e23bf3b521541d29d4ea0377c475059 profile=memory,time