A: An HPC cluster is a technical high performance computing cluster with workloads distributed to many "compute nodes". The cluster works as a single entity or system on specific tasks by performing compute-intensive operations on large sets of data. The cluster appears as one system to the systems outside the cluster. All communication to instances outside the cluster is performed by dedicated "head nodes" only. All compute nodes in the cluster have the same configuration.

A: An HPC workload performs compute-intensive operations on large sets of data. It can be a third party or custom application.

Many IT infrastructure services are not considered to be HPC workloads. For example, the following would not be HPC workloads: database, web application, file/print, backup, firewalls, DNS, Samba, Squid, DHCP and directory servers. High availability clusters or load balancer machines are also not considered HPC clusters.

A: Each initial purchase must be for at least 4 compute node subscriptions. Individual subscriptions can be purchased subsequently.

A: A Red Hat Enterprise Linux for HPC Compute Nodes subscription shares the same technology as a standard Red Hat Enterprise Linux server. The products share the same kernel, compiler tool chain, utilities, management tools, etc.

A: By placing a minimum cluster size and limiting the workloads for which Red Hat must provide maintenance and support, Red Hat is able to eliminate costs and pass on savings to customers that are deploying highly scalable HPC solutions. So services supplied with a Red Hat Enterprise Linux for HPC Compute Nodes subscription are only provided for configurations running HPC workloads.

A: Yes. Red Hat's compute node subscription provides maintenance and support for systems that have the same configuration. Head nodes, which must run Red Hat Enterprise Linux or Red Hat Enterprise Linux Advanced Platform, are not required to have the same configuration as the compute nodes.

A: Yes. This makes it much easier for customers to manage large clusters.

A: All standard Red Hat Enterprise Linux IHV/OEM hardware certifications apply to the HPC compute node product. A number of HPC-related ISV applications have also been certified. Please refer to your application vendor for details.

A: These two terms generally refer to the same thing:

HPC: High Performance Computing
HPCC: High Performance Computing Cluster

A: Contact sales on the web or call 1-866-273-3428 x45555.