The History of CQU's HPC facilities

In early 2002, CQUniversity had a small compute farm funded by a local faculty for computational research applications.  In 2006, CQU and QUT, with contributing funds from QPSF (Queensland Parallel Supercomputing Foundation – now known as QCIF), successfully partnered to win a national competitive "Linkage Infrastructure Equipment and Facilities (LIEF)" Grant.  This grant enabled both QUT and CQU to upgrade its High Performance Computing (HPC) facilities.  With this funding CQUniversity purchased and commissioned a significant SGI HPC System in 2007, which was affectionately known as the "Einstein HPC cluster".  This system provided significant compute and storage facilities to all CQUniversity academics, researchers and research students, until it was decommissioned in November 2012.

In 2012, CQUniversity successfully applied to the Queensland Cyber Infrastructure Foundation (QCIF) to supply matching fund of $200,000 for the "Upgrade of HPC facilities project".  Funds of $200,000 from the Office of Research and the Information Technology Directorate ($100,000 each) and QCIF matching funds lead to the purchase of CQUniversity third generation HPC facilities, in which was known as the Isaac Newton HPC Cluster.

In December 2017 and January 2018, CQUniversity conducted an upgrade to its HPC facility and whilst not a complete replacement, many of the core components including storage and compute was replaced. This fourth generation HPC facility is known as the Marie Curie HPC Cluster and like the previous HPC systems, this facility is available to all Central Queensland University academics, researchers and research students.  This system has been designed to support future growth in advanced computing utilisation.  The newest HPC system at CQ University allows significant computing capabilities that provides access to large CPU, memory and/or data storage resources.  Additionally, the HPC facilities utilizes a "Job Scheduler", thus allowing users to submit a significant number of jobs to be processed.  The CQUniversity HPC facility provides researchers access to unprecedented local research computing infrastructure.

The table below provides details of current and previous CQUniversity's HPC Systems:

Current Systems

 

OVERVIEW
Commissioning Date

Initial installation - January 2018

Additional upgrade occurred in October 2018

Additional upgrade occurred in December 2019

COMPUTE NODES
Number of Compute Nodes20
CPU Sockets40
Cores528
GPU4
Total Memory7.296 TB
Theoretical Performance63.65 Tflops
SHARED STORAGE
Disk Capacity320TB Raw
OTHER

Max Power

(Not including old nodes)

TBD

Isaac Newton Cluster (Decommissioned)

 

OVERVIEW
Commissioning Date21st of November 2012
COMPUTE NODES
Number of Compute Nodes31
CPU Sockets64
Cores544
GPU2
Total Memory4.35TB
Disk62TB Raw
Theoretical Performance12.35 Tflops
SHARED STORAGE
Disk Capacity240TB Raw
OTHER
Max Power21 KW

Decommissioned Systems

Einstein Cluster (Decommissioned)

OVERVIEW
Commisioning DateNovember 2007
Decommisioning Date16th of November 2012
COMPUTE NODES
Number of Compute Nodes20
CPU Sockets40
Cores160
GPUNone
Total Memory320GB
Disk10TB Raw
Theoretical Performance~1.16 Tflops
SHARED STORAGE
Disk Capacity6 TB Raw
OTHER
Max Power 

Colony Cluster (Decommissioned)

 

OVERVIEW
Commisioning DateEarly 2002
Decommisioning Date21st December 2007
COMPUTE NODES
Number of Compute Nodes84
CPU Sockets84
Cores84
Total Memory42GB
Disk1.38TB
Theoretical Performance84 GFlops (or 0.084 TFlops)
SHARED STORAGE
Disk Capacity138GB Raw
OTHER
Max PowerUnknown