CSIRO seeks $1.5 million supercomputer upgrade

The search is on for supercomputer implementation partners

Australia’s national scientific research agency, the CSIRO, is searching for a “suitably qualified and experienced” technology service provider to supply, install, and maintain a new Advanced Technology Cluster (ATC) at its Pawsey Supercomputing Centre in Western Australia.

The CSIRO is after an entirely new technology system that is to be located at the Pawsey Centre in Perth – the proposed procurement is for a three-year contract with a fixed budget of $1.5 million.

This includes hardware, software licensing, maintenance and support requirements, installation, and commissioning costs.

The Perth-based facility is used by the radio astronomy research community in the region, along with high-end researchers in other areas of computational science, including nanotechnology and biotechnology.

The centre’s resources are used for real-time processing of data for the astronomical research, storage and analysis of data from instruments at the Murchison Radioastronomy Observatory and other places of research, and scientific simulation at petascale computational speeds.

Among the centre’s computer systems are two Cray supercomputers, one of which is a Cray XC40 petascale system, named ‘Magnus,’ which is among the top 500 ranking such systems in the world and currently the most powerful public research supercomputer in the Southern Hemisphere.

As reported by ARN, the supercomputer reached this milestone in 2014 following a major upgrade.

According to request for tender published on 14 September, the facility is looking for an implementation partner that can provide an ATC to replace a decommissioned high-powered system located at the University of Western Australia’s Perth Campus.

The proposed ATC has two technology components, and providers competing in the tender process are expected to provide individual bids for each component in additional to a combined bid.

Each of the two technology components of the proposed system is to be split across the stated budget, with approximately $750,000 to be allocated to each one. The compute nodes in the first component of the system are to include at least one Xeon Phi processor.

Meanwhile, the second component of the system needs to include at least two GPUs with a Pascal architecture, and at least two host processors with either a Power or an Intel x86 64 bit architecture, according to the tender documents.

The installation, configuration, and acceptance of the system is expected to be completed before the end of April next year, and the centre is calling on tenderers to provide assurance that they are not aware of any evidence of exploitation in their supply chain.

The tender offer closes on 26 October.

Back to top