A five-person UCCS research team led by Brandon Runnels, assistant professor of mechanical and aerospace engineering, will address the regional needs for high-performance computing power through a National Science Foundation Major Research Instrumentation grant. The High-Performance Computing Cluster for Computational Science in Southern Colorado with acquire, deploy and maintain INCLINE, to provide much-needed computational resources to the UCCS campus and the southern Colorado scientific and academic communities.
Runnels will be joined by Qing Yi, associate professor of computer science; Byeong Lee, assistant professor of electrical and computer engineering; Amanda Morgenstern, assistant professor of chemistry; and Matt Quinlan, assistant professor of mechanical engineering.
“The size and power of INCLINE will bridge the growing gap between workstation-level machines and TOP500 supercomputers,” Runnels said. “This will allow researchers to test and leverage code scalability on a high-performance computing, or HPC, platform to expedite result processing, and to gain expertise on a local HPC environment. The work will yield insight into biomedical applications, such as microbubble drug delivery and bone fracture, military applications such as additively manufactured energetics, civil applications including improved structural materials, as well as educational and outreach opportunities.”
The NSF is providing $435,228 of the total $625,000 cost for the project.
INCLINE’s state-of-the-art hardware will be designed to support a broad range of high performance scientific applications, ranging from compiler design to computational fluid dynamics. The instrument will contain:
- 15 compute nodes, each with two 64-core AMD CPUs for a total of 128 cores and a 256 gigabyte registered dual in-line memory module, to facilitate standard massively parallel computations, such as computational solid mechanics and fluid dynamics.
- Two high memory nodes, identical to the compute nodes but with a two terabyte registered dual in-line memory module, that will be used primarily for parallel visualization.
- Two graphic processing unit nodes, each with a single NVidia V100 16GB GPU along with two 32 core processors and 512 gigabytes of registered dual in-line memory. The nodes will allow GPU acceleration on computational physics and machine learning projects and will be used in conjunction with central processing unit nodes to test optimal load balancing on heterogeneous architectures.
All nodes will be connected using InfiniBand high speed interconnects to minimize delay for communication-intense applications. A high speed SCRATCH storage file system will minimize input/output delay for applications with large output file sizes and requirements.
The estimated peak performance of INCLINE is approximately 90 teraflops, or 90 trillion floating-point calculations each second. The Slurm queueing system will be used to manage accounts and allocations across the diverse user base. The robust design of this instrument will allow it to fill the growing need for a local HPC research facility at UCCS and in southern Colorado and will facilitate the training of the next generation of computational scientists.