Storage and Compute
Hyalite Compute & Storage Cluster
The HYALITE Research Computing Cluster was installed in December, 2014. It is a small cluster of 65 (XEON) nodes for 2120 HT cores and has 620 TB of Lustre storage. HYALITE is able to churn out about 18 tera-flops of raw computing power.
The RCI file storage clusters offer large-scale data storage at a very competitive price. Since the storage systems are on campus, researchers can keep their data within the network of the University so as to not burden themselves (or the university) with off-campus data bandwidth costs.
Our primary storage cluster is comprised of high-availability, dual-redundant Supermicro servers, providing 290 TB (usable) of dynamic file storage, managed by Intel Enterprise Edition Lustre. This storage cluster was recently increased from 140 TB. Our secondary stroage cluster is comprised of dual-redundant, high-availability Dell servers that provide 48 TB (usable) of dynamic file storage, managed by the Gluster Storage System. Both of these storage clusters are highly scalable and will grow according to demand.
The RCI Storage Cloud platform offers a Globus Enpoint that researchers can use to transfer files from nearly any machine using Globus. Globus was developed to enable grid computing: by connecting computing resources, the data can be freed from its initial source, and made portable—even if it’s huge.
For more information please email firstname.lastname@example.org.
The RCI supplies computational resources to aid researchers in performing computationally intense operations. We currently have a AMD 64-core server with 512Gb of RAM for large computation purposes , an Intel 24-core machine with 124GB of RAM and an Intel 16-core server with 64GB of RAM. In addition, we act as a liaison between MSU and cloud service providers like Amazon and Microsoft via the Pacific Northwest Gigapop and Internet2. We also have two XSEDE campus champions with access to startup allocations for XSEDE HPC resources and can help researchers with deployment.