Progress in our procurements

To support the growing need from researchers for HPC and storage data the ANS2018 project will put in place two new HPC systems as well as an expansion of NIRD and the Service Platform.


HPC Machine C1

The upcoming new HPC-Machine, provisionally called C1, is currently in its final stages of the procurement process. Pre-qualifications were completed before the summer holidays and tenders submitted by the deadline in August. It is expected that the procurement process will be completed by end of November and  the delivery and installation finalized in Q2 2019. Following this, a 3 month approval period will commence to ensure that the system is  ready for production.


HPC Machine B1

The larger parallel HPC-Machine B1 has had an offset time schedule compared to  the C1 machine. This was done to cope with two large procurement processes. The  prequalification process was completed in August with only one qualified vendor. A single vendor in a competition such as this is obviously not ideal, thus a decision on how to proceed was thoroughly discussed  with the university partners and the Sigma2 board. 

Additionally, and as background material to these discussions, a vendor survey followed by an evaluation was conducted  to identify why  only one qualified vendor emerged. The conclusion of the evaluation and discussion was to cancel the competition with the intent to re-release it with some modifications made to the specifications. The new competition was published 17 October  and the plan is to complete the procurement process for B1 by Q1 2019.

Both of these machines will be hosted by NTNU, and preparations of the locations are ongoing to make sure that the housing facilities are ready and according to specifications.


NIRD and Service Platform

A second expansion of the NIRD storage facility is currently being planned, as well as an expansion of the Service Platform. For the NIRD, mainly storage volume will be the focus this time and the long term strategy is to reach approximately 11 PB replicated storage by end of 2019. For the Service Platform, the expansion includes adding one site (NTNU), expanding the number of worker nodes, and finally providing GPU capability.

More information: