Skip to main content

Node Buy-In Program

About the Program


Boise State’s Node Buy-in Program affords researchers a low-cost, flexible way to own computing resources without having to  purchase and maintain standalone systems.

Buy-in nodes are integrated into the shared cluster, and owners are given priority access to their equipment and can scale-out and use other resources available on the cluster. Excess capacity is returned to the pool for general, shared use. and in exchange for making idle compute cycles on their purchased nodes available, researchers pay no additional costs for managing their hardware or software or for using additional shared resources.

Participation in the Buy-in Program provides researchers with many benefits:

  • Favorable hardware pricing
  • Priority access to owned compute nodes with shared access to Research Computing cluster
  • The use of the shared cluster’s resilient, high-performance, parallel file-system for accessing their project’s files
  • The use of an Infiniband high-speed interconnect for larger parallel compute jobs
  • Professional staff management of the hardware, software and security environment

Equipment Services

Equipment purchased under the Buy-in Program will be integrated into Boise State’s compute cluster. Buy-in owners will have access to their home directory (backed up), project (both backed up and not backed up) shared storage space. All equipment purchased under this program that is charged to a sponsored project must be essential, allowable, and allocable. If the equipment is to be used for multiple sponsored projects and/or for non-sponsored research, an appropriate allocation of costs must be made among the funding sources, including departmental funds.

Equipment purchased under the Buy-in Program has a 5-year lifetime and will come with a 5-year, next business day hardware warranty. At the beginning of the 5th year, Research Computing will contact the researcher to discuss available options. If a researcher chooses not to upgrade compute nodes by the end of the 5th year, the node will be retired and Research Computing will recycle it using the existing salvage process.

If you are interested in participating in the Buy-in Program or have more questions, please contact Research Computing at to work on the following process:

  • Discuss the requirements
  • Set up project, accounts, and associated home directory and project disk space on the current shared system so work can begin immediately if needed
  • For Buy-In, establish access and priority policies that will apply to the researcher’s Buy-in hardware
  • Start the purchase process for their Buy-In hardware
  • Once Buy-In equipment arrives, Research Computing will place it into production, and the researcher will have immediate access

The following describes node buy-in services:

  • Installation: Compute nodes will be installed within a Shared Computing Cluster rack  and will be integrated into the shared file-system for the cluster.
  • Networking: By default an Infiniband IB connection is provided.
    Infrastructure: Power and cooling will be provided without charge.
  • Hardware maintenance: Five years of maintenance is included with purchase. We will provide coordination of hardware maintenance and repair with the vendor.
  • System administration: Your hardware will be integrated into the existing shared Linux cluster. We will be responsible for installing, configuring, and maintaining all system software.
  • Monitoring: We will provide automated monitoring of system status, hardware and software with real-time alerts, as well as regular and active monitoring of system security. We will respond to critical problems within 4 business hours. (See Support below)
  • File systems: Buy-in storage will be integrated into the cluster’s shared file system and allocated to the owner’s project(s). The shared file-system is accessible from all nodes, including Buy-in nodes, within the cluster.
  • Batch queues. Your nodes will participate in the batch queuing system for the cluster. (See Priority Access below)
  • Priority access:  Your nodes will be included in a priority queue for your use only, and will also be part of a shortq, with a 12-hour wall clock. This short queue includes nodes that have priority in other queues, as well as regular compute nodes. When a job is submitted to the shortq, it is with the knowledge that it will always be preempted by your queue if you call for your nodes.  The scheduler will then automatically requeue that job to other available nodes.
  • Internet server restriction: Due to security concerns, we can not allow internet servers (web, FTP, etc.) to
  • run on the cluster nodes. If you have such needs please speak with RCS staff to see how we can best accommodate your requirements.
  • Database restriction: The configuration of the nodes prohibits them from running database servers.
  • Hours of operation: RCS staff are available to resolve problems and provide assistance during normal business hours.
  • Support: Support is available though our normal support channels such as sending email to
  • Uptime: RCS schedules bi-annual maintenance outages when necessary.
  • Maintenance, update, and system changes: Emergency maintenance may occur with little or no advanced notice. Normal maintenance and system changes will be announced at least one week in advance. Any non-critical changes requiring downtime on a login/head node will take place during scheduled maintenance.    For compute nodes, we will first disable scheduling of jobs on that node and wait until all running jobs complete. Maintenance/changes will then be implemented during the maintenance window to allow the node(s) to be placed back in service as soon as possible.


Several standard configurations of compute nodes are available at specially negotiated prices and may be requested at any time. Please contact to discuss ordering requirements. The estimated cost per node for the Buy-In Program includes the cost of the node plus peripherals of ~$850 (cables and Bright licensing for 5 years), and networking infrastructure costs of ~$1,000. The cost of the node depends on specifications and can range as shown below (while configuration can vary, see Borah node specifications for examples here

  • Node Cost (~$12,500 standard CPU to ~$25,000+ for GPU or HIMEM, depending on specifications)

Purchasing Process

After consulting with the Research Computing Services technical staff, the researcher or the grant/departmental administrator should send an email to and and be prepared to discuss buying process and provide the information listed below:


Include email, departmental address and phone number for:

  • Person Requesting Buy-in
  • Fund Approver (must be able to approve shopping carts and funding sources indicated below):
  • Back-up Approver (in case primary person is away):


All funding sources that should be charged: (one or more of the following)

  • Fund
  • Department
  • Cost Center
  • Supplemental (if grant funded)
  • Project Future (if grant funded)

If multiple funding sources are being used, please supply us with a percentage or dollar breakdown of how costs should be distributed

In almost all circumstances orders will be considered a capital expense.

Grant-Based Funding: For grant-based funding sources, please coordinate with the the sponsored project administrator, academic department business manager, and associate director of research computing to ensure the purchasing process complies with uniform guidance and that the purchase is essential, allowable, and allocable. Note that the networking/infrastructure costs must be paid for with funds other than grant funds and the researcher must pay for the equipment directly.

Also, for grant-based funding sources please indicate when the funding will expire. We need to ensure that the funds can be invoiced in time.