Promise and challenge of high-performance computing, with examples from molecular modelling

Area of Science:

• Computational Science and Engineering
• Computer Science
• Applied Mathematics

Background:

• Computational modeling is a cornerstone of modern scientific inquiry.
• Rapid advancements in computing technology have led to significant increases in computer speed.
• These advances necessitate changes in computer architecture, posing new challenges.

Purpose of the Study:

• To discuss the challenges in utilizing high-performance computing (HPC) for critical scientific and engineering problems.
• To explore the necessary adaptations in software and theoretical areas to leverage HPC advancements.
• To highlight the importance of multidisciplinary collaboration in addressing these challenges.

Main Methods:

• Discussion of architectural changes in high-performance computing.
• Analysis of software requirements for new computer architectures.
• Review of theoretical advancements needed for effective computational science.
• Case study of NWCHEM, a computational chemistry code for parallel supercomputers.

Main Results:

• Realizing the benefits of HPC requires substantial software revision and theoretical progress.
• Close collaboration between computational scientists, computer scientists, and applied mathematicians is crucial.
• The NWCHEM code exemplifies successful multidisciplinary development for parallel architectures.

Conclusions:

• Overcoming HPC challenges demands a unified approach across scientific and engineering disciplines.
• Continued advancements in computational science depend on adapting to evolving computer architectures.
• Interdisciplinary teamwork is essential for pushing the boundaries of scientific problem-solving with HPC.