HPC research

Using FPGAs to model the atmosphere

Author: Nick Brown
Posted: 11 Dec 2019 | 15:54

The Met Office relies on some of the world’s most computationally intensive codes and works to very tight time constraints. It is important to explore and understand any technology that can potentially accelerate its codes, ultimately modelling the atmosphere and forecasting the weather more rapidly.

Field Programmable Gate Arrays (FPGAs) provide a large number of configurable logic blocks sitting within a sea of configurable interconnect. It has recently become easier for developers to convert their algorithms to configure these fundamental components and so execute their HPC codes in hardware rather than software. This has significant potential benefits for both performance and energy usage, but as FPGAs are so different from CPUs or GPUs, a key challenge is how we design our algorithms to leverage them.

EPCC’s ARM system: comparing the performance of MPI implementations

Author: Nick Brown
Posted: 9 Dec 2019 | 12:48

MVAPICH is a high performance implementation of MPI. It is specialised for InfiniBand, Omni-Path, Ethernet/iWARP, and RoCE communication technologies, but people generally use the default module loaded on their system. This is important because, as HPC programmers, we often optimise our codes but overlook the potential performance gains of better choice of MPI implementation.

Workshop review: 'Reducing Barriers to HPC Adoption for SMEs'

Author: William Lucas
Posted: 25 Nov 2019 | 12:45

On 14 November 2019, EPCC hosted the workshop ‘Reducing Barriers to HPC Adoption for SMEs’. Held in the University of Edinburgh’s Bayes Centre, the one-day meeting’s aim was to increase awareness of the benefits of high-performance computing (HPC) use among European small and medium-sized enterprises (SMEs), in particular of the gateway offered to them through the HPC-Europa3 programme. In total twenty-six representatives from ten SMEs and three HPC-Europa3 centres attended.

Integrating 3D remeshing in the iceberg-calving model of Elmer/Ice at St Andrews

Author: Guest blogger
Posted: 19 Nov 2019 | 17:10

Eef van Dongen visited the University of St Andrews from 23 September–11 October 2019 through the HPC-Europa3 transnational access programme. In this post she describes her work on improving iceberg-calving simulation techniques.

Hi! My name is Eef van Dongen. I am a PhD student in glaciology at ETH Zurich. This autumn I visited Dr Joe Todd in the research group of Professor Doug Benn based at the University of St Andrews for three weeks with HPC-Europa3. My research focuses on iceberg-calving – the breaking off of icebergs at the edge of glaciers into the sea – which is responsible for approximately half of the ice mass loss of the Greenland ice sheet.

Modelling chemical reactions in (explicit) solvents

Author: Guest blogger
Posted: 18 Nov 2019 | 16:02

Arthur’s Seat in Edinburgh. St Andrews should be straight ahead.

Arthur’s Seat in Edinburgh. St Andrews should be straight ahead.

Nicolas Sieffert was an HPC-Europa3 visitor to the University of St Andrews in summer 2018 and again in 2019. In this blog post he describes his two trips to Scotland.

Hi there! I am Nicolas Sieffert, associate professor in computational chemistry at the University Grenoble Alpes (France). I visited the UK for nine weeks under the HPC-Europa3 transnational access programme and was hosted by Prof. M. Bühl at the University of St Andrews (UK), with HPC resources and technical assistance provided by 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.

NEXTGenIO: the end is just the beginning

Author: Michele Weiland
Posted: 7 Nov 2019 | 14:55

After four years of hard work, the NEXTGenIO project has now come to an end. It has been an extremely enjoyable and successful collaboration with a dedicated group of HPC users, software and tools developers, and hardware providers from across Europe.

Precision persistent programming

Author: Adrian Jackson
Posted: 30 Oct 2019 | 12:48

Targeted Performance

Optane DIMM

Blog post updated 8th November 2019 to add Figure 6 highlighting PMDK vs fsdax performance for a range of node counts.

Following on from the recent blog post on our initial performance experiences when using byte-addressable persistent memory (B-APM) in the form of Intel's Optane DCPMM memory modules for data storage and access within compute nodes, we have been exploring performance and programming such memory beyond simple filesystem functionality.

For our previous performance results we used what is known as a fsdax (Filesystem Direct Access) filesystem, which enables bypassing the operating system (O/S) page cache and associated extra memory copies for I/O operations. We were using an ext4 filesystem on fsdax, although ext2 and xfs filesystems are also supported.

Exploring the binding mechanism of glycomimetics to galectins through simulations

Author: Guest blogger
Posted: 21 Oct 2019 | 11:13

Domenica Capasso was an HPC-Europa visitor to University College London from 12 June to 11 August 2019. In this blog article she describes her work and time in London.

In June this year, I returned to London and I spent two months at the Department of Chemistry at UCL, funded by HPC-Europa and hosted by Professor Francesco Luigi Gervasio. My project concerns the galectins, β-D-galactoside binding proteins with important implications for tumorigenesis, inflammatory response and autoimmune disorders. During my first visit, we created a virtual library of candidate compounds, containing two saccharide residues with a bridging sulphur/selenium atom. These derivatives, targeting galectins, were obtained using structure-based drug design approaches and are synthesised and tested in my Institute in Naples.

Global or local - which is best?

Author: Adrian Jackson
Posted: 9 Oct 2019 | 17:30

Selfish performance

Sharing of resources has challenges for the performance and scaling of large parallel applications. In the NEXTGenIO project we have been focusing specifically on I/O and data management/storage costs, working from the realisation that current filesystems will struggle to efficiently load and store data from millions of processes or tasks all requesting different data sets or bits of information.

Pages