Part of the work carried out at the Institute is assembling the wheat genome. Thanks to Allinea MAP, which shows developers exactly where code is losing performance, the Earlham team was able to speed up the run time of applications by 50 percent. Allinea's tools helped to make sure its research was delivered much faster, bringing the Institute's goals closer to fruition.
The Institute is developing a method of mapping-by-sequencing using three SGI supercomputers, an SGI UV2000 with three partitions, and two new SGI UV300s - all with Intel processors. Scientific Computing Specialist, Luis Yanes, supports the work of the researchers by finding the right tools to work on the systems used and he has turned to Allinea MAP, part of the Allinea Forge suite of tools, and Allinea Performance Reports, to speed up research time.
"Before, we were using open source software to analyze performance issues", Luis Yanes explained. "There were problems though as this was a very slow, manual process. Open source tools aren't as robust and didn't clearly show us where the problem was or how to modify the algorithms to our advantage. MAP's integration with SLURM and PBS schedulers allowed us to easily queue the profiling jobs to use the busy servers."
Luis Yanes explained that the wheat genome application had previously spent too much time in synchronization. Allinea MAP allows a deep insight into memory usage and vectorization relating to the Intel processors - something the team found invaluable.
Wheat is one of the three major crop plants: find a high quality wheat genome and the world is on the way to increasing yield stability and vastly boosting crop production. With Allinea MAP enabling a 50 percent application speed up time, the Earlham Institute is able to continue its valuable research into solving world food issues at pace.