Thanks for visiting Scientific Computing World.

You're trying to access an editorial feature that is only available to logged in, registered users of Scientific Computing World. Registering is completely free, so why not sign up with us?

By registering, as well as being able to browse all content on the site without further interruption, you'll also have the option to receive our magazine (multiple times a year) and our email newsletters.

Numerical Algorithms Group aids study of egg shell formation

Share this on social media:

University of Warwick researchers Mark Rodger and David Quigley, in collaboration with colleagues at the University of Sheffield, have used molecular dynamics simulation and the HECToR supercomputer – a 12,000 core Cray XT4 – to solve a problem in eggshell formation. Protein ovocledidin-17 (OC-17) is found only in the mineral region of eggs and laboratory results showed that it appeared to influence the transformation of calcium carbonate into calcite crystals. Having been unclear as to how this process could be used to form an eggshell, the researchers created a model to show how the protein bound to a calcium carbonate surface.

In addition to the Cray supercomputers managed by the University of Edinburgh, the Research Councils’ HECToR service includes a comprehensive Computational Science and Engineering (CSE) support service provided by the Numerical Algorithms Group (NAG). One of NAG's HPC experts identified input and especially output (I/O) as the bottleneck for the simulation software, known as DL_POLY_3 – originating from Daresbury Laboratories.

A solution was provided for this research and NAG has gone on to optimise all the I/O routines, to assist other users of DL_POLY_3. This was done by reordering data to take advantage of modern file systems and then further improved by performing the I/O in parallel. This six months of work has resulted in the I/O for DL_POLY_3 being around 50 times faster on average.

Dr David Quigley from the Department of Physics and Centre for Scientific Computing, University of Warwick, said: ‘Prior to the I/O improvements, DL_POLY_3 was unable to make effective use of the parallel file system on HECToR, severely crippling the performance of our simulations. The new code has reduced the time taken to write a single snapshot from three minutes to less than half a second, resulting in an overall factor of 20 improvement in our net performance. Without this development, HECToR would have been effectively useless for our purposes.’