Gen X: ExCALIBUR working group on Exascale continuum mechanics through code generation.

点击翻译按钮获取中文摘要

Continuous physical processes pervade every aspect of our society,industry and the natural world. From the flow of air over an aircraftto the propagation of mobile phone signals, to the behaviour ofchemical components at every point of the manufacturing processes,continuum mechanics is at the heart of our industrial processes. Inmedicine, the electrical behaviour of the heart and brain, the flow ofblood and other fluids through the body, and the detection of disordersusing all manner of scanners and detectors are all continuum mechanicsprocesses. In the natural world, detecting and understanding themovement and composition of the Earth enable us to understandearthquakes and to hunt for valuable minerals, while advancedunderstanding of the complex interaction of fluids and electromagneticfields allows us to understand stars, the cosmos and our place in it.In all of these cases and many more beside, the mathematical equationsdescribing phenomena are known, but solutions very rarelyexist. Science and engineering are essentially dependent on computersimulation to understand any of these systems, and to design thedevices and processes which use them. Many of these phenomena are socomplex or have such a range of spatial scales that existing petascalecomputer systems are a limit on scientific advance. In addition, thereis a need to go beyond mere simulation to simulate the uncertainty inprocesses, find the optimal solution, or discover the multiplepossible outcomes of a system. The advent of exascale computingpresents the opportunity to address these limitations. However,increasing computational scale, increasingly complex simulationalgorithms, and the vast quantities of data produced by exascalecomputing will defeat not just existing simulation software, but alsoexisting ways of writing simulation software.Gen X is a project to establish the requirements for exascalesimulation software for continuum mechanics, and to provide a concreteway of achieving this capability within the next five years. The Gen Xapproach is to move beyond just writing code to a system of specialistsimulation languages which enable scientists and engineers to specifythe problem they want to solve and the algorithms they want by writingmathematics, the language of science. The actual code will beautomatically generated by specialist compilers rather thanhand-written. Rather than an algorithm developer writing a paper abouttheir new development and hoping that simulation scientists will findthe time to code it up for their specific problem, the algorithm willbe encoded in a domain specific language and implemented in itscompiler. The simulation scientist will then be able to access thealgorithm directly without recoding.At exascale, writing all the simulation outputs to disk for lateranalysis is impossible. Instead, simulation data must be processed,analysed and visualised as the simulation is conducted, and only theresults stored for later use. Gen X will provide mathematicallanguages for this process which will enable the scientist or engineerto concisely specify the analysis to be performed, and to haveconfidence that the resulting calculations will be both efficient andcorrect.By enabling scientists and engineers to work at a higher mathematicallevel while also accessing more sophisticated algorithms andhardware-specific implementations than previously possible, Gen X willmake simulation science both more capable and more productive. In thismanner, Gen X is essential to realising the potential of exascalecomputing while also making the most efficient use of researchresources.