


ACM Challenges
The following is a selection of problems which continue to drive
the development of the field at large. Climate prediction
 Physical challenge: complicated  coupled
atmosphereoceanice dynamics plus atmospheric chemistry and
evolution of the biosphere
 Mathematical challenge: cannot accurately predict the
weather in 100 years, so algorithms have to be validated
statistically: what is the climate in 100 years?
 Computational Challenge: required computer power is
huge, massively parallel algorithms necessary
Molecular dynamics
 Computational Challenge: simulation of large systems
over very long times (relative to the smallest relevant nearatomic
subsystems)
 Physical challenge: Quantum vs. classical dynamics;
accounting for the environment
 Mathematical challenge: How to characterize molecular
states and sample possible transitions between them well;
coarsegraining and reduced models
Tomography
 Mathematical challenge: Reconstruct a threedimensional
representation of the body only from multiple twodimensional
projections  an underdetermined problem
 Physical challenge: model the properties of the Xray
detectors as best as possible
 Computational Challenge: be fast, want images in
realtime
Image sources: Screenshot of ClimatePrediction.net v5.15
running under Boinc v5.4.9
from Wikipdedia;
chest CT scan with lung metastasis from
Wikipedia; simulation of proteinprotein docking by Jacobs
faculty member by Martin
Zacharias.

