Increased computational capacity for Australian researchers

Australian researchers will receive increased computer capacity, with the CSIRO and National Computational Infrastructure (NCI) set to implement improved systems.

XENON Systems has been selected to supply a Lenovo NeXtScale system for NCI, while CSIRO has begun searching for the next generation of supercomputer to replace its current Bragg accelerator cluster.

As requirements for access to high-performance computing (HPC) capabilities continue to increase, the Lenovo system will help NCI to meet this demand by providing a 40% increase in capacity. It will be integrated with Raijin’s fast filesystems, with key specifications including 22,792 Intel Xeon Broadwell 2690v4 processors; 144 TB of memory, including ten 1 TB nodes; and one Mellanox EDR 100 Gbps Infiniband interconnect, configured as a ‘fat tree’.

The NCI upgrades and 2017 replacement of the oldest of the global parallel filesystems will be supported by the federal government 2015–16 Agility Fund for the National Collaborative Research Infrastructure Strategy (NCRIS).

The new facility will enter production in January 2017 and will extend vital advanced computational services for national research and innovation through NCI. 

Meanwhile, Bragg’s replacement at CSIRO will be capable of ‘petaflop’ speeds, significantly exceeding the existing computer’s performance. It will boost CSIRO’s HPC capability and is expected to rank highly on the Green500.

CSIRO’s acting deputy chief information officer for scientific computing, Angus Macoustra, said this replacement computer would be essential to maintaining CSIRO’s ability to solve many of the most important emerging science problems.

“It’s an integral part of our strategy working alongside national peak computing facilities to build Australian HPC capacity to accelerate great science and innovation,” Macoustra said.

The cluster will power a new generation of groundbreaking scientific research, including data analysis, modelling and simulation in a variety of science domains, such as biophysics, material science, molecular modelling, marine science, geochemical modelling, computational fluid dynamics and, more recently, artificial intelligence and data analytics.

The tender for the new machine is calling for a heterogeneous system, combining traditional central processing units with coprocessors to accelerate both the machine’s performance and energy efficiency.

The successful bidder will be asked to deliver and support the system for three years within a $4 million proposed budget.

Image credit: ©FreeImages.com/Ariel da Silva Parreira