Simulations and Supercomputers
At the forefront of science is the exploitation of cutting edge computer hardware, such as supercomputers, to simulate real life systems and potential systems of interest. With more computational power, scientists can simulate bigger systems, more detailed systems, and they can simulate longer timescales. In addition, with that extra power scientists can compute outcomes faster, making predictions of important events before they occur.
To gain anything of value from a simulation, compromises must be made, even on supercomputers: detail can be simplified, timescales can be reduced, or smaller systems considered. The insights that you can gain from a simulation based on one theoretical / mathematical description, over a particular space and time scale (for example, a 1 second simulation of a molecule) are often different to what you gain from a simulation with a different set of such scales (for example, a 1 nanosecond simulation of a protein). Scientists have found that, by connecting simulations at different scales, you can improve the output and gain further insight. These are called multiscale simulations.
Making compromises in the space and time of simulations is fundamental in making them feasible to run on modern computer hardware. However, because of this, any output from a simulated system may be compromised due to the assumptions and approximations made.
It is critical that we can trust the results of our simulations, because simulations are relied on for important matters that affect our lives, and will be even more so in the future. This has created a big push for validating, verifying, and quantifying the uncertainty in simulations.
In the VECMA project, we’re looking to make multiscale simulations on cutting edge supercomputers Verified, Validated, and have their Uncertainty Quantified. VECMA stands for Verified Exascale Computing for Multiscale Applications. The term ‘exascale’, represents the computing power of the next generation of supercomputers. A 1 ExaFLOP supercomputer will be capable of a quintillion (1, followed by eighteen 0s) calculations per second.
The VECMA project is arming scientists with the tools needed to run verifiable multiscale simulations on exascale computers. VECMA is developing a toolkit, called the VECMA Toolkit, that allows scientists to verify their multiscale simulations with ease.
Through the power of the VECMA toolkit, a number of computer applications have already benefited:
Climate – Climate models use simulations to show the drivers of climate, including atmosphere, oceans, land surface and ice. They can be used to give projections of future climate, which need to be verifiable if the conclusions are to be relied upon. VECMA is exploring simulations of the atmosphere and oceans, both key components of the climate system, which are currently limited because of computational limitations, but can be unlocked with the power of multiscale simulations.
Migration – When conflicts break out, refugees inevitably migrate away from the combat areas. VECMA aims to use verified mutliscale simulations in order to forecast refugee movements, guiding decisions on where to provide food and infrastructure, acquire approximate refugee population estimates in regions where existing data is incomplete, to prioritise resources to the most important areas, investigate how border closures and other policy decisions are likely to affect the movements and destinations of refugees, to provide policy decision-makers with evidence that could support more effective policy and reduce unintended consequences.
Materials – Computational modelling and simulation has become increasingly common in materials research, assisting experimental laboratory work with quantitative predictions from simulations. By doing so, time and money can be saved by steering the design of new materials with particular desired properties. This requires exploring the mechanical properties of a material at several length and time scales with verified multiscale simulations.
Fusion – Nuclear fusion potentially provides a carbon free solution to the provision of base load electricity, without geo-political complications.
In a fusion process, two lighter atomic nuclei (an atom stripped of electrons) combine to form a heavier nucleus, while releasing energy. VECMA aims to use verified mutliscale simulations to understand the mechanisms of heat and particle transport in present and future fusion devices, which is important because the transport plays a key role in determining the size of a future fusion power reactor, and hence the cost of electricity.
Biomedicine – Computational biomedicine uses simulations to make patient-specific medical decisions. To use such simulations in the clinic, decisions need to be reliable, quick, and cost effective. Verified multiscale simulations can achieve this. VECMA aims to use verified multiscale simulations to aid in the decision-making of drug prescriptions, but simulating how the drug interacts with a virtual versions of your proteins. VECMA also aims to simulate how stents will behave when placed in virtual versions of your arteries.