Translational Research Institute Implements SGI HPC

With 1PB on disk and 3PB on tape

This is a Press Release edited by on 2013.06.12
AddThis Social Bookmark Button

Silicon Graphics
International Corporation
(SGI) announced that Translational Research Institute (TRI) has
selected SGI to provide a big data HPC solution powered by SGI UV 2000 shared
memory platform, SGI Rackable clusters and SGI InfiniteStorage to accelerate
results at its research centre.


This new facility, which represents four leading medical research institutes,
will focus on advanced treatments and therapies for common and serious disease
such as cancers, diabetes, inflammatory diseases, HIV, malaria, bone and joint
diseases and obesity. The institute is destined to be the largest biomedical
research institute in the Southern Hemisphere.

Researchers will now have access to the necessary technology and facilities in
one location, which will positively impact productivity through the rate at
which work is processed and scientific results are achieved. SGI's compute and
storage solution provides more than 2,200 SGI Rackable compute cores, 256 cores
and 4TB of memory UV 2000 and more than 1PB of InfiniteStoragestorage. Up to 3PB of historical and inactive data will be
stored on tape and available via SGI's DMF system.

The combination of SGI Rackable scale out clusters, UV 2000 shared memory for
large 'in memory' application requirements along with high performance
storage offered the flexibility of compute, storage platforms and software that
TRI needed to tackle their big data problems. This solution will complement the
massive amounts of data that high resolution gene sequencers, microscopes and
associated laboratory equipment generate. The SGI solution will assist in
increasing productivity and accelerating the time to discovery of new
treatments, subsequent commercialisation and significant patents.

TRI COO Dr. Kate Johnston highlights the rapid advances seen in R&D over
the last decade, signaling that scientific research is fast becoming an
exercise in handling enormous data sets.

"Scientists are essentially looking
for the veritable 'needle in a haystack' in amongst this data,
" said Johnston. "Without the correct
technology to do this it could take researchers years to find their 'needle.'
The SGI High Performance cluster provides our researchers with the computing
power they need in order to both analyse and appropriately store their large
experimental datasets.

SGI has a long and positive reputation of providing high performance
infrastructure to enable Australian research projects. Dr. Eng Lim Goh, SGI's
SVP and CTO, speaks about how important it is for SGI to be a part of this
pioneering work.

"The storage, analytical and
computational needs of TRI are substantial and unique,
" said Goh.
"SGI with our big data solutions are
proud to be a part of TRI as they translate their discoveries to advances in
patient care.

TRI's collective expertise across common and serious diseases such as cancers,
diabetes, inflammatory diseases, HIV, malaria, bone and joint diseases, obesity
and children's health research will have a direct impact on improved public
health and enhanced preventative treatments for people worldwide. The Institute
brings four of Australia's leading research facilities together with the aim of
translating the findings of basic biomedical research into better patient