DDaT22008 UKRI - BGS HPC Cluster

UK Research and Innovation (UKRI) consists of 9 Councils, including 7 research councils, Innovate UK and Research England. British Geological Survey (BGS) is a component body of Natural Environment Research Council (NERC).

One of the growth areas in the Geosciences is modelling at ever increasing resolution, with the production of geomagnetic models as one example of this. High Performance Computing (HPC) allows rapid and timely computation of global magnetic field models, used for academic, public good and commercial purposes. BGS is at the forefront of development of these models, which require millions of data points solving for thousands of parameters and could not be calculated without a HPC. BGS also uses HPC to deliver an increasing number of its key public tasks. These range from hydrological modelling and seismic hazards to complex high-resolution 3D and 4D geological models.

To fulfil these needs, BGS has operated an HPC for many years. The current cluster was bought in 2014.

Due to the time since a HPC cluster was last sourced, BGS are conducting some market research to get feedback on their specification, and understand available technological options, pricing mechanisms and potential supply chain issues. A route to market has not yet been established.

We're holding a pre-market supplier engagement session on Thursday 17th March 2022 at 11am for an hour.

IMPORTANT

To ensure participation, Expression of Interest emails must be received by ddatprocurement@uksbs.co.uk no later than midday (UK time) on Wednesday 16th March 2022.

In the message subject field please state: "EoI DDaT22008 Pre-market engagement", and provide up to two names and email addresses of your attending representatives. A drafted specification and details to join the session will be issued.

All attending suppliers must familiarise themselves with the spec in advance of the session as we are seeking expertise and feedback from the market. We will also be requesting input regarding pricing mechanisms.