Scientist – EuroEXA Project



1. Position information

Vacancy No.: VN17-08
Grade: A2
Job Ref. No.: STF-PL/17-08
Department: Research
Section: Earth System Modelling
Reports to: Head of Section
Closing Date: 10 August 2017

2. About ECMWF *

ECMWF is both a research institute and a 24/7 operational service, producing and disseminating numerical weather predictions to its Member States. This data is fully available to the national meteorological services in the Member States. The Centre also offers a catalogue of forecast data that can be purchased by businesses worldwide and other commercial customers. The supercomputer facility (and associated data archive) at ECMWF is one of the largest of its type in Europe and Member States can use 25% of its capacity for their own purposes.

For details, see www.ecmwf.int/.

ECMWF uses its high-performance computing (HPC) facility to produce a time-critical twice-daily global numerical weather forecast. The operational forecasting system and research & development are currently run on two Cray XC-40 clusters totalling over 250,000 Intel Broadwell compute cores each. Given the expected need for increasing the Earth system model resolution and complexity, enhanced parallelism in work flows and computer code are required, as may be the exploitation of emerging hardware architectures such as low-power multicore CPUs combined with FPGAs.

3. Summary of the role

In 2014, ECMWF embarked on a ten-year research programme on HPC Scalability. This programme is funded partly by ECMWF’s Member States and aims to achieve exascale numerical weather prediction capability by 2025.

As part of the Scalability Programme, ECMWF is engaged in partnerships with its Member State organisations, HPC centres and vendors to prepare for the exascale challenge, in terms of both computing and data handling. The EuroEXA project is funded by the European Commission under its FETHPC-2016 programme. EuroEXA aims to install a world-class ARM-based HPC platform, on which the computational study of exascale problems from climate, energy and life sciences domains will be showcased.

For this project, ECMWF is opening a post to optimise performance of the ECMWF weather forecasting model on ARM CPUs and to investigate the use of FPGAs for specific aspects of this model.

4. Main duties and key responsibilities

The successful candidate will perform scientific and technical research and development in the following areas, and document work according to the requirements of the EuroEXA project:

  • ? Investigating the impact of data layout, OpenMP usage, loop structure on the IFS performance on ARM hardware
  • ? Investigating the use of a Domain-Specific-Language (DSL) to implement parts of the ECMWF IFS model. One such DSL could be CLAW1, developed at MeteoSwiss
  • ? Modifying and tuning an existing CPU DSL toolchain for improved performance on an ARM system
  • ? Performing energy efficiency diagnostics on ARM and x86 systems

5. Personal attributes

  • ? Excellent analytical and problem-solving skills with a proactive approach, together with an interest in identifying, investigating and solving technical challenges
  • ? Dedication and enthusiasm to work in a small team
  • ? Excellent interpersonal and communication skills
  • ? Ability to work under pressure

6. Qualifications and experience required

Education

  • A university degree, or equivalent, in a discipline related to computer science, meteorology, physics, mathematics or engineering is required.
  • A PhD is desirable but not essential

Experience

  • Experience in developing codes on high-performance computing facilities, including distributed systems, involving several levels of parallelism. Experience with accelerators such as GPUs or FPGAs, and/or with the use of a DSL, would be an advantage.
  • Experience with developing large scientific codes.
  • Experience working in projects and teams governed by tight schedules would be advantageous.

Knowledge and skills

  • Good knowledge of object-oriented programming and of a high-level programming language such as C/C++, Fortran. Knowledge of Python would be useful.
  • Strong background in UNIX systems (and/or Linux).
  • Good understanding of parallel programming (e.g. MPI and OpenMP).

Language

  • Candidates must be able to work effectively in English and interviews will be conducted in English.
  • A good knowledge of one of the Centre’s other working languages (French or German) would be an advantage.

7. Other information

Grade remuneration

The successful candidate will be recruited at the A2 grade, according to the scales of the Co-ordinated Organisations and the annual basic salary will be Ł56,487net of tax. This position is assigned to the employment category STF-PL as defined in the Staff Regulations.

Full details of salary scales and allowances are available on the ECMWF website at www.ecmwf.int/en/about/jobs/jobs-ecmwf, including the Centre’s Staff Regulations regarding the terms and conditions of employment.

Starting date: As soon as possible
Length of contract: Up to 36 months but not beyond 28 February 2021.
Location: The position will be based in the Reading area, in Berkshire, United Kingdom.

8. How to apply

Please apply by completing the online application form available at www.ecmwf.int/en/about/jobs/jobs-ecmwf.

ECMWF has an Equal Opportunities Policy and applications from all suitably qualified candidates are welcome.

Staff are usually recruited from among nationals of the following Member States and Co-operating States:

Austria, Belgium, Bulgaria, Croatia, Czech Republic, Denmark, Estonia, Finland, former Yugoslav Republic of Macedonia, France, Hungary, Germany, Greece, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, Montenegro, the Netherlands, Norway, Portugal, Romania, Serbia, Slovakia, Slovenia, Spain, Sweden, Switzerland, Turkey and the United Kingdom.

Staff from other countries may be considered in exceptional cases

published: 25 July 2017     Please mention EARTHWORKS when responding to this advertisement.