The ecological modeling lab of University of Toronto has selected Huawei’s high-performance computing (HPC) solution for its cutting-edge research.
Researchers of the ecological modeling lab, led by Dr. George Arhonditsis, at the University of Toronto will use the Huawei high-performance computing (HPC) fat node solution to perform four-dimensional (space and time) simulations associated with integrated watershed-receiving water body models.
Huawei HPC fat node solution is based on the Huawei FusionServer 8100 V3 8-socket mission critical server. With each server node supporting up to 144 compute cores, the system supports a capacity of up to 12 TB memory and 192TB storage. The system also supports 40G/100G high-speed Ethernet.
All of these system advantages are 5 times higher than the requirements of computing-intensive applications from the ecological modeling lab and cover expansion requirement for long term business development.
The University of Toronto is a world-renowned public research university at the forefront with academic research in the physical, electronic, and medical fields. The Department of Physical & Environmental Sciences, which is located at the University of Toronto Scarborough and is leading the ecological modeling lab, has a great need for computational power and RAM memory that will be used to support the implementation of Bayesian inference techniques and other uncertainty analysis and parameter sampling (Markov Chain Monte Carlo, Latin Hypercube) methods for cutting-edge researches. Therefore significant memory and storage capacity per HPC node are highly desirable.
Huawei offers a wide range of HPC fat node solutions based on the KunLun 9032 & 9016, FusionServer 8100 V3, and FusionServer 5885 V3 mission critical servers to satisfy the computing demands of various industries and applications. Huawei’s HPC fat node solution not only meets the performance indicator requirements of complex computing models, but also incorporates Huawei Reliability, Availability, and Serviceability (RAS) 2.0 technologies.
By supporting all-round system status monitoring and proactive alarm reporting, the solution drives system reliability to 99.999%. The hardware has a modular design and supports online maintenance and replacement of critical parts such as CPUs and DIMMs, increasing the reliability and availability of the HPC fat nodes.
Huawei provides HPC solution capabilities with optimized integrated software and hardware based on in-depth understanding about the HPC requirements of education and scientific research sectors.
Huawei has built customized HPC platforms for universities and research institutes including the University of Warsaw (Poland), the Newcastle University (UK), and Uppsala University (Sweden). Huawei has also built joint innovation centers with strategic partners such as the Poznan Supercomputing and Networking Center (PCSS) of Poland, maintaining a long-term stable technical partnership.