Scientists increasingly need extensive computing power to solve complex problems in physics, mechanics and dynamics. Delft High Performance Computing Centre (DHPC) will deploy the infrastructure (hardware, software and staff) for TU Delft that is capable of complex analysis and modelling for researchers. At the same time we will provide Bachelor, Master and PhD students with hands-on experience using the tools they will need in their careers.
Both high-performance simulations and high-performance data science are evolving rapidly and the combination of these techniques will lead to completely new insights into science and engineering, an increase in innovation, and the training of high-performance computing engineers for the future.
Due to the rapidly evolving hardware and tools for numerical simulations, HPC has significantly changed the way fundamental research is conducted at universities. Simulations not only replace experiments, but also add very valuable fundamental insights. We see the results in all kinds of disciplines, such as materials science, fluid dynamics, quantum mechanics, design optimization, big data mining and artificial intelligence.
Capacity, speed and functionality
DHPC offers optimal capacity and speed through a balanced price/performance choice of processor cores, memory and a high-speed connection between the nodes and a fast, parallel storage system.
DHPC offers a more heterogeneous solution instead of a uniform HPC design that will benefit the great diversity of applications used at TU Delft. Specific applications can make use of “worker nodes” that, for example, have more memory available, offer advantages through available GPU accelerators, can use faster storage (sub)systems or offer interactive access for tasks that are unsuitable for batch processing. All this with the goal of providing Delft with the necessary expansion and improvement of current and future HPC resources.
Reliability and availability
The hardware and software combination must be extremely reliable and generate reproducible results. Off course the cluster must be available 24/7.
The HPC cluster will be connected to TU Delft’s existing infrastructure for login, authentication, storage and licenses. Researchers and students will have access to the HPC cluster with their NetID.
The user experiences with the running HPC’s on campus and in use at SURFsara combined with the input from DCSE research groups form a ‘wish list’ for the new cluster configuration:
- CPU and GPU capacity;
- A fast parallel storage system for temporary work data;
- Fast data transport between compute nodes and memory and between the nodes themselves and the storage system;
- Software that must be able to work on the cluster.