EPCC

UrgentHPC SC19 workshop next week: see you in Denver!

Author: Nick Brown
Posted: 12 Nov 2019 | 11:11

Here in EPCC we lead a work package of the VESTEC EU FET project which is working on the fusion of real-time data and HPC for urgent decision-making for disaster response. While HPC has a long history of simulating disasters, what’s missing to support emergency, urgent, decision-making is fast, real-time acquisition of data and the ability to guarantee time constraints.

Online LAMMPS training for ARCHER

Author: Julien Sindt
Posted: 31 Oct 2019 | 15:56

In October I ran an online training course teaching current and potential ARCHER users how to run LAMMPS. LAMMPS is a widely-used open-source molecular dynamics (MD) software developed at the National Sandia Laboratories, and is optimised to run well on HPC facilities. It consistently ranks in the top 10 most used programs on ARCHER. Although the course was run on ARCHER and with ARCHER users in mind, I tried to teach users how to use LAMMPS on any platform.

Fun and Games in Historic Melrose

Author: Lorna Smith
Posted: 23 Oct 2019 | 14:48

Set in the heart of the Scottish Borders and home to Melrose Abbey and Melrose RFC, a visit to Melrose is always a pleasure. Organised by St Mary's School, Bang Goes the Borders (21st September 2019) is a science festival aimed at families. Entirely free, the festival has loads of workshops and drop in sessions.  The ARCHER team and Wee Archie​ were sat alongside workshops and activities around lego robots, apple juice making, plaque busting, bees, birds, and much, much more.

ARCHER Data Transfer Guide now updated

Author: Caoimhin Laoide-Kemp
Posted: 13 Sep 2019 | 10:39

As ARCHER and the UK Research Data Facility (RDF) are both coming to the end of their service, many users will have to transfer their data to their own machines. Because of this, the ARCHER Data Transfer Guide has been updated to give more precise and up-to-date information about best practices for transferring data between ARCHER, the RDF, and remote machines.

Optimising sparse matrix-vector products at EPCC

Author: Guest blogger
Posted: 15 Aug 2019 | 11:35

Edoardo Coronado visited EPCC from 18th May–15th August 2019 under the HPC-Europa3 programme. 

Hi! My name is Edoardo Coronado. I am a Mexican PhD student in Technologies of Information at the "Centro Singular de Investigación en Tecnoloxías Intelixentes" (CiTIUS) at the "Universidad de Santiago de Compostela" in Spain. I work on the optimisation of iterative solvers on accelerator devices. I have spent three months at EPCC at the University of Edinburgh under the HPC-Europa3 transnational access programme with Mr Adrian Jackson acting as my host.

ARCHER Training report Summer 2019

Author: Clair Barrass
Posted: 7 Aug 2019 | 15:59

ARCHER Training report Summer 2019

So, it's been another busy end to the academic year and summer training season for the ARCHER Training team.

Since June we have managed to cram in no less than twelve training courses, in locations from London to Leeds, and Oxford to EPCC here in Edinburgh.

Promoting Science on Computers to Schools - Workshop Event

Author: Clair Barrass
Posted: 16 Apr 2019 | 11:27

EPCC and EXDCI-2Funded by EXDCI-2, EPCC is organising a 2-day workshop in Edinburgh on 8-9 July 2019 to present ideas on how science and engineering can be supported from Raspberry Pis to large scale supercomputers. We are looking to investigate how to introduce these topics to school-age audiences including linking to existing school curricula. This will include examples of existing education materials and activities and looking ahead to current plans.

Proof-driven queries to preserve patient privacy

Author: Mike Jackson
Posted: 4 Mar 2019 | 09:42

StethoscopeIn our role as members of the Research Engineering Group of the Alan Turing Institute, Anna Roubickova and I worked with Efi Tsamoura and Benjamin Spencer (Department of Computer Science at the University of Oxford) on PDQ, a proof-driven query planner that has great potential within the realm of data science for medical research. 

What is MPI “nonblocking” for? Correctness and performance

Author: Daniel Holmes
Posted: 27 Feb 2019 | 15:53

The MPI Standard states that nonblocking communication operations can be used to “improve performance… by overlapping communication with computation”. This is an important performance optimisation in many parallel programs, especially when scaling up to large systems with lots of inter-process communication.

However, nonblocking operations can also help with making a code correct – without introducing additional dependencies that can degrade performance.

Fortissimo Marketplace: open for business

Author: Mark Sawyer
Posted: 15 Jan 2019 | 11:52

 

The Fortissimo 2 project ended on 31 December 2018. Together with its predecessor (the plain old 'Fortissimo project') it has helped over 100 SMEs and mid-caps to run experiments that demonstrate the effectiveness of providing HPC services using a business model derived from cloud computing, thereby making it much lower risk for small companies to use HPC. 

Pages