After twelve months of new management and eight months after a complete revamp of the organisational structure, positive results are beginning to emerge from the supercomputing centre (1). The autonomous unit of ETH Zurich has set itself ambitious goals, and is tenaciously pursuing them. Since the start of 2004 the centre has been run on the basis of a concrete working plan that aims to continuously strengthen scientific computing.

The aim behind the plan is to provide better and stronger support to fundamental research and scientific knowledge in Switzerland. First steps and successes of this plan were presented to a parliamentary delegation by Marie-Christine Sawley, together with Monica Duca-Widmer from the ETH Board and Professor Ulrich W. Suter, ETH Vice-president Research, and afterwards to the media.

Initial difficulties overcome

"We now have a clearer vision of where we want to go and can chart the route to get there," said Sawley after the media conference. She said this with allusion to the initial difficulties the centre was faced with and the subsequent numerous changes in management. It seems that these problems are a thing of the past. Over recent months the new management of CSCS had achieved basic progress that has already begun to nurture cantonal and national research.

The pride of the computing centre: the IBM Regatta with a capacity of 1.3 TFlops. large

A development, says Ulrich W. Suter, which he puts down to the new organisational planning of the computing centre. Earlier problems were, basically, organisational and not personal. "With the inclusion of the ETH Board we've now found the appropriate model for the centre – and with Marie-Christine Sawley the right head," says Suter and adds, "Now something has emerged that can show itself, at national and international levels."

Mid-term funding assured

"In order for the CSCS to survive against the background of global competition, much remains to be done," says Monica Duca-Widmer. As far as the question of financing is concerned, she is counting on political support. Funding is assured until 2007 and, consequently, the on-going development of the centre. "That's the main thing for time being," she says, optimistically.

Taking stock in Manno: Marie-Christine Sawley reporting on her first year as head of CSCS. large

For Sawley, the development and establishment of CSCS as an international brand also plays a part. International rankings, such as the list of the Top 500 Supercomputers (2), naturally play a role here. In the main she sees such grading systems as positive, driving, as they do, competition, but also because they provide the only possibility of comparison in the industry. But Sawley also sees their limits, because such rankings cannot measure which scientific content the supercomputers serve.

Three strategic linchpins

CSCS wants to combine high-speed, high-performance technology with useful research. Right from the start the new leadership banked on rapid developments connected by three strategic linchpins. The first of these is the renewal and further development of supercomputing capacities. With investments of around 20 million CHF until 2007 CSCS wants to achieve an increase of computing power by a factor of 5, which means a top performance level of 6 billion computing operations per second. At the same time a continuous supply of computing power must be guaranteed to cover the requirements of research.

The second strategic point is the strengthening of technical and scientific collaboration. New structures to support scientific co-operation need to be put into place here. Important partners at the national level are, e.g. the Paul Scherrer Institute and the Swiss Institute for Bio-Informatics. In an international context the new European programme, EGEE (Enabling Grids for E-Science in Europe) (3) is important.

The third strategic linchpin is the launch of a programme for a Swiss national grid (4). CSCS plays an important role in this project as a competence platform, co-ordinator and as a national resource. At present it is hoped that this supercomputing network will lead to seminal research results in the areas of particle physics and the life sciences.

People are the basis

In the short term personnel development is a priority. Until the end of the year the team is to be strengthened by 50 per cent with the addition of ten engineers and computer specialists. Personnel planning foresees that this increase will allow the development of competencies, a flat hierarchy and a subsidiary internal leadership structure. Sawley is certain of one thing: in the long run only a highly motivated team can work successfully. This is why she wants all staff to be firmly embedded in the centre. "If one feels oneself to be part of something, one automatically gives of one's best," she sums up.