Our HPC Resources

Our HPC facilities at Newcastle University have different access and funding models. Some of our facilities operate under a split free-access / funded-access model where a certain amount of resources are reserved for users without funding or who are working on activity to pump-prime further projects.

As a result, whilst everyone is able to access the HPC facilities, in order to access the larger pool of resources in several of our facilities, you will need to allocate funding using the standard University TRAC system for HPC services.

HPC Facilities

Newcastle University has offered a number of HPC facilities to our research community. Their status is shown below.

Facility Status Description
Rocket
Decommissioning Rocket is an HPC service used by Newcastle University. It was introduced in 2017 as the first institution-wide HPC facility, and is open to all staff and students, regardless of funding. Rocket is primarily used for CPU-based compute, although has a small amount of GPU resources. There are no priority queues or resources implemented on Rocket - all users and projects have equal access to all available nodes, with everyone being subject to a 'fair use' policy.
Comet
Operational Comet is the replacement for the previous Rocket facility. It builds upon the capabilities of Rocket by offering a significantly increased set of CPU resources, increased memory limits and additional GPU resources. The Comet facility also introduces, for the first time at Newcastle, container based job support and the ability to run compute-intensive graphical applications through Open OnDemand. Comet was brought online for the Newcastle University HPC community in late 2025 through an academic-lead design and procurement process.
Compute Resources

If you understand the type of resource your computation requires then you can select it from the list below and generate a costing for your project or research activity.

Contributions & Costings

You can also jump straight to the costing tool if needed

Note: You must be logged in to access the costing tool.

Facility Status Resource Type Nodes Unfunded Use CPU CPU Arch Cores/Node Cores/Total Frequency RAM/Node GPU/Node GPU/Total
Rocket Decommissioning Power.a 1 100% 2x IBM Power 9 Power 32 32 2600 - 3090 MHz 256 GB 4x Nvidia Tesla V100 (16 GB) 4
Rocket Decommissioning Standard.a 110 100% 2x Intel Xeon E5-2699 v4 x86_64 44 4840 2200 - 3600 MHz 128 GB - -
Rocket Decommissioning Medium.a 6 100% 2x Intel Xeon E5-2699 v4 x86_64 44 264 2200 - 3600 MHz 512 GB - -
Rocket Decommissioning Large.a 4 100% 2x Intel Xeon E5-2699 v4 x86_64 44 176 2200 - 3600 MHz 512 GB - -
Rocket Decommissioning XL.a 2 100% 4x Intel Xeon E7-4830 v4 x86_64 56 112 2000 - 2200 MHz 1536 GB - -
Comet Operational Standard.b 36 20% 2x AMD EPYC 9745 x86_64 256 9216 2400 - 3700 MHz 1152 GB - -
Comet Operational Large.b 10 25% 2x AMD EPYC 9745 x86_64 256 2560 2400 - 3700 MHz 1536 GB - -
Comet Operational Standard.Lowlatency 4 25% 2x AMD EPYC 9745 x86_64 256 1024 2400 - 3700 MHz 1152 GB - -
Comet Operational GPU-S 4 25% 2x AMD EPYC 9745 x86_64 256 1024 2400 - 3700 MHz 768 GB 8x Nvidia L40S (48 GB) 32
Comet Operational GPU-L 1 0% 2x AMD EPYC 9745 x86_64 256 256 2400 - 3700 MHz 768 GB 4x Nvidia H100 (96 GB) 4
Rocket Decommissioning Dell-GPU.a 1 100% 2x Intel Xeon Gold 6242 x86_64 32 32 2800 - 3900 MHz 1536 GB 1x Nvidia Quadro RTX 6000/8000 (48 GB) 1