Chapter 5
Industrial Applications
on Supercomputing
in Germany
Michael M. Resch and Andreas Wierse
5.1 History 64
5.1.1 Industrial Take-Up 64
5.1.2 Supercomputing in Research 65
5.2 Current Supercomputing Service Provider Landscape 66
5.2.1 Supercomputing Service Provider Landscape
for Research 66
5.2.2 Supercomputing Service Provider Landscape
for Industry 68
5.3 German Supercomputing Ecosystem 70
5.4 German Supercomputing Industry 71
5.5 Case Studies and Business Cases 72
5.5.1 Public–Private Partnership 72
5.5.2 Porsche: A Success Story 73
5.5.3 General Business Case 74
5.5.4 Small- and Medium-Sized Enterprises 74 Large-Scale Power Plant Simulation 75 Simulation of Production in Clean Rooms 76
5.5.5 Stimulation Plans for Adoption of Supercomputing by
the German Industry and SMEs 76
5.6 Conclusion 78
64 Industrial Applications of High-Performance Computing
Computing has a long history in Germany and is rooted deeply both
in research and in industry. Two of the pioneers of supercomputing are
German. David Hilbert (1862–1943), a prominent mathematician, gave an
overview talk about the key problems of the coming twentieth century
during the Worlds Fair in Paris in 1900. Among the topics he mentioned
were the key questions for the numerical solution of partial dierential
equations that were the basis for many of todays simulations both in
research and industry. Hilbert was also an outstanding teacher. In the
years aer his talk in Paris, a number of excellent scientists joined his
institute in Göttingen. Among them were Richard Courant, who was his
PhD student, and John von Neumann with whom he worked on quantum
mechanics. e second outstanding supercomputing pioneer was Konrad
Zuse (1910–1995). In 1939, he built his rst computer (Z1), and, in 1941, he
built Z3, the worlds rst programmable computer.
e history and career of both men highlight the strengths and weak-
nesses of German supercomputing. David Hilbert is a representative of the
German mathematical school that very early focused on applic ation-driven
problems and has continued to do so until today. However, between 1933
and 1945 many of the most brilliant scientists had to leave Nazi Germany—a
brain drain that damaged Germanys science substantially. Konrad Zuses
invention was not recognized as being relevant by Germanys public
authorities, then both the Z1 and the Z3 were destroyed during the war.
Although a number of German electronics companies, such as Telefunken
and Siemens, actively worked in the eld, German hardware industry
never again achieved a relevant worldwide position.
However, both David Hilbert and Konrad Zuse still stand out in the
history of supercomputing in Germany. First and foremost, industrial use
is connected closely to scientic progress. German industry and German
research organizations traditionally cultivate a close collaboration. is is
also seen in the use of supercomputing.
5.1.1 Industrial Take-Up
Early industrial adopters of supercomputing were found mainly in three
dierent elds. e importance of supercomputing became evident in
nuclear power plant engineering. To the extent that nuclear power became
important for energy supply in Germany in the 1960s, companies started
to use supercomputers for the simulation of critical processes. e use
of large-scale systems included the development of soware. Nuclear

Get Industrial Applications of High-Performance Computing now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.