In the last post, we learned about persistent workloads hosted in the Arbutus research computing cloud and the contributions these project make to their respective fields. On the other side of the advanced research computing coin are the compute workloads. Recall, compute workloads are usually ephemeral processes which are memory- and/or storage-intensive and tend to start and finish quickly. These jobs can make use of more cores than persistent projects tend to, up to 1400 cores, and usually take several hours to several days to complete.
Hermes
The Hermes High Performance Computing (HPC) cluster is a compute project that enables scheduling and execution of general purpose compute jobs on the Arbutus system. Hermes is running on the cloud with 1.2 PB of storage and is dedicated to UVic research. Using this dedicated batch system, users can submit and manage their jobs, almost as if these systems were in their offices, customizing the computing power to meet the needs of their projects. Many jobs running on Hermes use statistical modeling to unlock critical research insights. One user of Hermes, Juan Aldana, uses statistical modeling and bioinformatics data from the Western Red Cedar tree to determine mechanisms by which some of the tree populations are susceptible to a particularly damaging type of fungus. This research supports our understanding of the genetic factors that impact the multi-billion-dollar forestry industry of British Columbia. Another user of Hermes is Juan Francisco Chavez Ramirez from the Business department. Juan is working with the Research Computing Services (RCS) team to determine which types of rhetoric are likely to garner success for project proposals under review by the National Energy Board. The National Energy Board has reviewed major projects including the Transmountain pipeline and plays an integral role in the approval of these kinds of projects. Juan’s project has taken the transcripts of the stakeholder review meetings and analyzed them to distill effective language and types of arguments common among successful projects. This analysis can inform a user’s ability to present information in a way that has proven to be effective, and increase the chances of success for their project as it is reviewed by this influential board.
ONC
Ocean Networks Canada’s (ONC) research projects are also on the Arbutus cloud system. This important project simulates tsunamis and earthquakes to support predictive modeling for natural disasters. It provides information to support effective emergency planning and disaster management efforts. The ONC project also hosts the Neuron Project, a machine learning initiative that is able to pick out noteworthy data points from vast data stores, think hours of video footage streaming from the bottom of the ocean, and compile it in to a sort of ‘highlight reel’ for researchers to review anomalies. This project helps to save the valuable time of researchers, limiting their review to only the novel results captured by cameras and freeing their time to commit to high-value activities.
CANUE
Finally, the Canadian Urban Environmental (CANUE) Research Consortium project uses Arbutus persistent resources to coordinate health-related data of Canadians with geographic information, providing insights in to the impact of environment on our health. CANUE has two servers, one running a Windows server for ARCGIS analysis and another running UBUNTU. The latter server processes health-related data and hosts the platform’s website, while the ARCGIS server provides different layers of geographical information for analysis. Using machine learning algorithms, the CANUE platform analyses the connection that complex environmental factors, such as nearby traffic, noise, pollution and access to greenspaces, have on the health outcomes of Canadians. Due to privacy regulations for health-related data, it is imperative that this analysis remain in Canada. Arbutus makes it possible for key connections to be discovered while abiding by these strict regulations. This project can inform city planning and space usage and aims to contribute to the development of social policy that better supports positive health outcomes for Canadians.
To learn more about ARC in Canada, explore the links throughout this article and check out these websites: