HPC aPPliCations: BiomediCal
continues to buy additional sequencers,
it will be able to scale the system up to a
maximum bandwidth of 200GB per second
by adding additional nodes and scaling
up their storage. This means that TGen
will not face bottlenecks as it continues to
expand its sequencing capacity.
The difficulties involved in obtaining
such high performance become apparent
from the fact that TGen’s storage system
serves a compute cluster with 4,680 nodes,
each of which has multiple cores. When
it comes to analysing the terabytes of raw
data produced by a single sequencer run,
the massive parallelism of the compute
cluster must be matched by the speed of
the single storage solution in order to avoid
potential performance bottlenecks.
Fuchs states that each high-performance
storage system must be designed from
➤ genetic data must be acquired, stored and the start if it is to reach an adequate
processed. ‘The users do not want to have a performance level, and that these
The human genome project spurred a
room full of expensive sequencing
considerations become the driving force
drive within research institutions around the behind all design choices. The design of
world to find ever faster ways of achieving
machines standing around doing
the RAID system, the development of
accurate genetic sequencing. Now, two nothing, as they wait for the the control algorithms, and the physical
decades after the project began, there
storage solution to catch up.
enclosures themselves are all engineered
exist many research groups specialising for performance. ‘This approach gives
in sequencing genomes, with a view to
In addition, multiple disparate
us a tremendous performance advantage U58 130x213:Layout 1 21/09/2009 08:42 Page 1
learning more about the roles and effects of storage solutions scatter data over solutions based on a windows
individual genes within each person. The
over several systems, making it
kernel or similar,’ explains Fuchs. The
process of sequencing the genome has been system is designed to provide predictable
automated to a high level, and modern
more difficult to manage, and so
performance, even if failed disks are being
QUOTE SCW AND RECEIVE £100 OF DELEGATE PLACE
*
sequencers are able to complete the task a single consolidated system is rebuilt by the RAID controller in the *NOT TO BE USED IN CONJUNCTION WITH ANY OTHER OFFER.
in a fraction of the time it once took. So
preferable.’
background.
SMi present their 3rd Annual Conference…
the demands upon the sequencing centre’s As the demand for sequencing increases,
storage facilities have increased. ever more computing power is required
Electronic Laboratory Notebooks
A single sequencing run on a state-of- by the institutes performing the work
the-art machine produces 7TB of data senior product manager with Data Direct and subsequent analysis. Data centre
Wednesday 27th and Thursday 28th January 2010
per run. Of this 7TB, only 100-200GB Networks, a provider of customised storage costs represent a massive expense for
Crowne Plaza St James, London
of usable data may be kept afterwards, systems for HPC. ‘In addition, multiple such companies, in terms of floor space,
but with thousands of sequencing runs disparate storage solutions scatter data over cooling, and power supply (including
Our expert line-up of speakers includes:
performed every year, the demands on several systems, making it more difficult UPSs). Minimising the physical size of the
• Michael Kopach, Principal Research Scientist, Eli Lilly
storage soon add up. One run can take to manage, and so a single consolidated storage system allows research institutions • Letitia Fearon, Research Lead, ELN Technologies, Pfizer
Sponsored by
anything between a few hours and a few system is preferable.’ to manage their storage costs. The
• Omer Casher, Manager, Medical Imaging IT, GlaxoSmithKline
days. Multiple sequencers are routinely TGen (Translational Genomics Research solution used by TGen stores a petabyte
• Antonio Gomez, Senior Scientist, Information Management, Johnson & Johnson
linked to a single storage system, meaning Institute) is a non-profit organisation based in a standard server cabinet. ‘All of these
• Carla Marchioro, Director & Site Head, Verona Analytical Chemistry,
that the storage system must be fast enough in Arizona. It is expanding its line-up of infrastructure considerations may not have
GlaxoSmithKline
to take in data as quickly as the sequencers sequencers to include five next-generation that whole life science “flare” to them,’
can produce it in order to ensure that ABI Solid systems, and more sequencers Fuchs acknowledges, ‘but in the end you
the sequencers are utilised to the limits will be added thereafter as demand need the right tools to get the results, and
PLUS TWO HALF-DAY WORKSHOPS
of their capabilities. ‘The users do not increases. It has installed a storage system, we’re providing those; they are the tools in
want to have a room full of expensive designed by Data Direct Networks, capable the background that make it happen.’
A ELNs – Concept to Reality
sequencing machines standing around of taking in 5GB per second, which is fast Perhaps the TV dialogue of ‘scalpel.
PM Tuesday January 26th 2010, Crowne Plaza St James, London
doing nothing, as they wait for the storage enough to handle the data output from the . . swab . . . supercomputer’ is not so far
www.smi-online.co.uk/2010eln2.asp
In association with Amphora
solution to catch up,’ says Igno Fuchs, current sequencers. As the organisation distant after all. Register online and receive full information on all of SMi’s conferences
B
Laboratory Integration – The Next Hurdle?
Alternatively contact Victoria Bull on Tel +44 (0) 20 7827 6158
36
AM Friday January 29th 2010, Crowne Plaza St James, London
SCIENTIFIC COMPUTING WORLD october/november 2009
www.scientific-computing.com
or email
vbull@smi-online.co.uk
In association with phaseFour Informatics
SCWOct09 pp32-37 Bio.indd 36 10/1/09 1:28:05 PM
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48 |
Page 49 |
Page 50 |
Page 51 |
Page 52 |
Page 53 |
Page 54 |
Page 55 |
Page 56