[Q-e-developers] next developers meeting

[Layla Martin-Samos]

Dear all

it is time to start organizing the next developers meeting. We think that a
good period could be just before, during or just after the Total Energy, to
be held in Trieste 10th-12th January 2013 (note that the Total Energy is
followed by a workshop that involves some of us, from the 14th to the
18th). Could you please send us an email with your preferred dates?

Thank you very much

Paolo and Layla

PS In attachment please find a short summary of the past developers meeting.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/developers/attachments/20121003/a5b36510/attachment.html>
-------------- next part --------------
QE developers meeting, Trieste 24 jan 2012
Participants:
1-Carlo Cavazzoni
2-Lorenzo Paulatto
3-Filippo Spiga
4-Fabio Affinito
5-Stefano de Gironcoli
6-Andrea Ferretti
7-Henry Lambert
8-Harry Fisher
9-Nicola Seriani
10-Ralph Gebauer
11-Andrea Dal Corso
12-Erica Vidal
13-Tone Kokalj
14-Paolo Umari
15-Riccardo Sabatini
16-Emine Kucukbenli
17-Alan Bizjak
18-Changru Ma
19-Marco Monni
20-Layla Martin-Samos

PG give an introductory presentation with the following topics
1) One year later: status of planned activities
	-QE foundation: established, no activity yet
	-Modularization and packaging: mostly done
	-Moka builder, new xml i/o: stalled
	-Binaries: volunteers sought
	-Maintainer for each package: volunteers sought
2) More ongoing or new activities
	-PP validation and verification: PSlibrary and AE calculations
	-New website
	-Massive parallelization (PRACE)
	-New developments
	-New Mailing lists/Forum
3) Reminder
	-Please subscribe to q-e-developers and q-e-commits
	-Never develop on old QE versions
	-use svn facilities
	-Read and Keep updated the developer man
---------------------------------------------
GPU by Filippo Spiga
	-Addusdens -> openmp 20X faster
	-vloc_psi -> serial 9X faster, parallel -> 4X faster
	-MAGMA lib (serial diagonalization) DSYGV and DSYGVX replaced by DSYGVD
	-full gpu porting 4X faster
	-porting stress and forces
	-The memory is an open issue
--------------------------------------------
SOLVENT model Oliviero Andreussi
	-Done but needs Cleaning
--------------------------------------------
QMMM by Changru Ma
	-Electrostatic coupling: ongoing 
	-Cleaning is needed
--------------------------------------------
EPW by Harry Fisher
	-http://epw.org.uk
	-Band gap renormalization
	-synchro with 5.0 pending: problems with symmetries
----------------------------------------------
AE calculation V&V by Marco Monni
	-LDA no spin polarization
	-SC FCC and BCC + molecules of the G1 set (55 molecules)
----------------------------------------------
HPC by Carlo Cavazzoni
	-BGQ architecture FERMI. Min jobs 2048 cores, 128 mpi
	-Cineca engaged to help porting for side codes
	-Organization of a special openmp training for QE and YAMBO developers?
	-Ongoing work on:
		-Linear Algebra
		-Matrix-Vector (H-psi) operations
		-Extension of multithreads on the whole code (openmp+tests with openACC)
---------------------------------------------
TDDFPT by Simon Binnie
	-EXX with TDDFPT: ongoing
	-PAW in EXX: to be done
--------------------------------------------
GWL by Paolo Umari
	-Collinear spin
	-fractional occupations
	-off-diagonal states
	-approx. treatment of semi-core states (on fock part included)
	-Analytic continuation
	-Contour integration
	-Openmp on blas and lapack
	-Ongoing: GPU porting
--------------------------------------------
Comments, ideas etc
	-LAMMPS tutorial, for promoting the use of QMMM? (R. Gebauer)
	-pw_export with the spin? (A. Ferretti)
	-Reduce the number of executables, in particular for PP?
	-General/portable drivers for running many DFT calculations?
	-Delete ifdef __PARA before mp calls
	-davcio_drho is a mess, possible changes?
	-Involving more the Irish community, CECAM workshop? (I. Girotto)