DDN Provides University of Tennessee’s SimCenter with Big Data Storage to Support Machine Learning and Data Analytics
DataDirect Networks (DDN®) today announced that The University of Tennessee, Chattanooga (UTC) has selected DDN’s GS14KX® parallel file system appliance with 1.1PB of storage to replace its aging big data storage system and to support a diversifying range of data-intensive research projects. The Center of Excellence in Applied Computational Science and Engineering (SimCenter) at UTC needed a big data storage solution that could scale easily to support growing research programs focused on computational fluid dynamics (CFD), machine learning, data analytics, smart cities and molecular biology. The DDN GS14KX is purpose built to address the comprehensive needs of HPC environments and manage huge data growth, enabling organisations, such as SimCenter, to scale their environments and take advantage of new, data-intensive research disciplines such as machine learning.
The SimCenter is a research incubator with both an inward-facing role to support the growth of innovative research at UTC and an outward-facing role to support and collaborate in making Chattanooga a world-class smart city. “The real-time data coming from smart cities is huge; every street will have data sets coming in about everything from small Bluetooth devices to autonomous vehicles,” said Anthony Skjellum, Director of SimCenter at UTC. “The scalability and quality of integration from DDN will allow us to implement a number of smart city test beds connecting into the new storage solution, with machine learning algorithms taking advantage of the near real-time data. We have a torrent of data so the ability to scale up capacity or performance was vital, and DDN was able to provide a clear path for growth.”
Since the start of the SimCenter in 2002, it has served as a hub in modeling, simulation, and high-performance computing for all colleges at UTC and is a research incubator that helps UTC build its innovative doctoral programs in computational science and computational engineering. Traditionally focused on CFD projects such as hypersonic flows, which typically have very large data sets, the SimCenter also supports research into energy and environment, manufacturing, urban systems and smart cities, and health and biological systems.
A number of users are taking advantage of GPU hardware in the SimCenter’s newest Dell EMC HPC cluster to speed up research. "DDN’s powerful big data storage platform will give us the ability to easily meet the diverse and dynamic demands of our current research programs and enable us to scale performance or capacity as we look to expand our research capabilities,” said Ethan Hereth, High Performance Computing Specialist at UTC. “The integration of DDN’s solution into our existing Dell EMC HPC infrastructure has been fabulous. As well as being larger than our previous storage system, the density is far better. We are only using a quarter of the rack space as before for the same capacity, which we could easily double with DDN by only adding new hard drives.”
The SimCenter needed to modernise and future-proof its storage to support the expanding research focus and enable researchers to access data quicker, meet the demands of applications with larger I/O bandwidth requirements, and support the collaborative smart city project. With DDN’s highly scalable storage solution, SimCenter is prepared to meet the growing demands of data-intensive applications and research, providing superior performance. As well as providing 8x better density than before, the solution will be filled entirely with self-encrypting hard drives (SEDs), which will enable UTC to access research grants, initiatives and projects that have requirements around encryption.
“DDN has a proven track record and has a very good reputation. They met the challenge and delivered a really good solution, providing everything that we required. We wanted a solution that was easily scalable and that would seamlessly integrate into our existing environment. The collaboration between Dell EMC and DDN provides an end-to-end HPC storage solution that fits into our infrastructure and adds tremendous value,” added Skjellum.
This article has been republished from materials provided by DataDirect Networks (DDN®). Note: material may have been edited for length and content. For further information, please contact the cited source.
Computer scientists at Carnegie Mellon University say neural networks and supervised machine learning techniques can efficiently characterize cells that have been studied using single cell RNA-sequencing (scRNA-seq). This finding could help researchers identify new cell subtypes and differentiate between healthy and diseased cells.