31 / 54
High-Performance Computing
High-Performance Computing or HPC, is the application of “supercomputers” to computational problems that are either too
large or would take too long to be completed on a standard computer. A desktop computer generally has a single processing
chip, commonly called a CPU. An HPC system, on the other hand, is essentially a network of nodes, each of which contains
one or more processing chips, as well as its own memory.
The HPC cluster in UAE University has been operational since 1st July 2017. The HPC nodes are interconnected via high-speed,
low-latency 10Gbps Ethernet and Infiniband switches. Below is our current HPC cluster hardware specifications:
Currently, most of our HPC users are from the College of Science running computationally intensive mathematical
and physics applications. The HPC CPU cores mean utilization currently stood at around 40% of its capacity with peak
utilization at around 70% of its capacity. We are expecting growing demands for HPC access in the coming months as
well as encouraging the whole research communityat UAE University to utilize the HPC system. We are planning to
expand the HPC hardware capacity as well as adding more scientific, research-related software to the HPC in the near
future.
»
»
45 teraflops theoretical performance
»
»
8TB total RAM
»
»
40TB shared storage
32 x CPU compute nodes each with:
2 x GPU compute nodes each with:
In summary, the capacity of our HPC cluster is as follows:
»
»
CPU: 2 x Intel Xeon E5-2697 v4 @ 2.30GHz (18 cores per-CPU; 36 cores total)
»
»
Memory: 256GB RAM
»
»
CPU: 2 x Intel Xeon E5-2695 v4 @ 2.10GHz (18 cores per-CPU; 36 cores total)
»
»
Memory: 256GB RAM
»
»
GPU: 2 x NVIDIA Tesla K80
29