Overview and additional information on services
RTO Research Computing (RC) provides High Performance Computing (HPC) facilities and related services. These services are focused on supporting HPC and are constrained to RC’s infrastructure and to be used to facilitate Research and academics. Research Computing does not provide services for ASU Administrative Use. RC does not support servers or any computing equipment outside of RC’s infrastructure.
HPC: Research Computing (RC) operates HPC supercomputers as a Core Service for research and academics. The use of the cluster(s) is free on a “fair share” basis to researchers, faculty, and staff. Students have access to the cluster through a sponsored research or faculty member. Use of the cluster and related systems and services is subject to agreement to an MOU that all users must agree to.
Purchases: RC will obtain quotes, and purchase transaction assistance for computing equipment that meets
RC’s definition of HPC grade equipment.
Installation, Configuration, and Maintenance: RC will install, configure, and maintain equipment that is to be included in RC’s infrastructure including compute nodes that are part of the RC’s computing cluster or computing hardware that is part of RC’s Virtual Machine (VM) infrastructure. RC does not support computing equipment that is not part of RC’s infrastructure.
Virtual Machines (VMs): RC will install computing nodes into RC’s VM infrastructure as long as the VMs are used for Research or academics are tied to RC’s HPC services, and fit within RC’s security posture for that area of infrastructure. RC will configure VMs for Research purposes and install a base OS if that base OS is a standard OS used by and supported by RC. RC may assist with the installation of other software and software-based services on VMs, at RC's discretion and with advance notification of any charges. This could include the installation of software such as database software or software development software, but the configuration, maintenance, and compliance of Operating Systems and Software on those systems (as noted in the definitions below, item 6) are the sole responsibility of the customer and will not be performed by RC. RC will provide for the conversion of the bare metal system to VMs upon procurement of services with four (4) hours of migration assistance included. RC will provide guidance on OS patching if the OS is an RC standard, and its utilization has not been customized by the customer to deviate from the Policies and Standards of ASU, KE, or Research Computing, otherwise, the customer is responsible for making sure the OS is patched and maintains an approved ASU security condition.
Co-located Equipment (colo): RC provides colo services to equipment that is RC grade and for Research or Academic purposes and is suitable for RC’s security posture. RC will install the equipment including networking. RC will install RC standard Operating Systems (OS), however the customer is responsible for any license costs for those. The customer is responsible for the maintenance and licensing of the server and all software, including the OS, installed on those systems. The customer is responsible for security and OS patches so that all parts of the systems meet ASU, KE, and RC security standards. The customer is expected to keep the equipment in a maintenance agreement. RC will assist in maintenance and repairs while it is under warranty or an acceptable maintenance contract. RC will not be responsible for any repairs or maintenance to equipment that is out of warranty. The customer recognizes that there is an expense borne by RC and ASU in providing rack space with power and cooling. RC may remove equipment if it becomes a security issue or inoperative to maintain a security posture and proper use of the expense of rack space, power, cooling, and associated services.
Storage: Data storage space is offered to the research and academic community. 100 GB is provided free to all HPC users. Additional project storage, for research and HPC academic use is available for additional fees as detailed on RC’s websites. Scratch storage is ultra-fast attached storage to the supercomputers for HPC-demanding jobs. Users are given storage on the “Home” system which is also attached and can be used for HPC computational needs. Retention policies are defined by RC’s documentation systems and are copied into this document for quick reference. However, the RC website maintains the current retention policies for all storage systems and will be the true policy.