Today’s high performance heterogeneous computing platforms offer highly advanced architectures that, if harnessed efficiently, can be used to power scientific discoveries as well as transform research with the help of computer modeling and simulations. Previously infeasible computational problems can now be solved with the aid of the increased computational power offered by these advanced hardware platforms.
However, in order to achieve large speedups, the core logic of many applications would need to exploit the different levels of parallelism offered by these platforms in order to make efficient use of hardware resources. Furthermore, doing so requires a detailed knowledge of the target platform and the parallel programming paradigms in optimizing the application logic, which is an area of active and ongoing research. The High Performance Computing (HPC) capability group aims to develop scalable and reusable solutions to enable developers to produce efficient and high performance applications. Building on these frameworks, the group looks to build high performance tools for scientific computing, as well as data analytics and machine learning.
Dr. Huynh Phung Huynh
Capability Group Manager
High Performance Computing