Brochure FUJITSU Supercomputer PRIMEHPC FX10
Copyright 2013 FUJITSU LIMITED http://www.fujitsu.com /global/services/solutions/tc/hpc/products/primehpc/
FUJITSU Supercomputer PRIMEHPC FX10
provides the ability to address these high mag nitude
problems by delivering over
23 Petaflops, a quantum leap in processing p e rformance.
Combining high performance, scalability, and
reliability with superior energy efficiency,
PRIMEHPC FX10 further improves on Fujitsu's
supercomputer technology employed in the "K
computer," which achieved the world's
top-ranked performance*¹ in November 2011.
All of the supercomputer's components—from
processo rs to middleware—have been
developed by Fujitsu, thereby delivering high
levels of reliability and operability. The system
can be scaled to meet customer needs, up to a
1,024 rack configuration achieving a
super-high speed of 23.2 PFLOPS.
Fujitsu has developed PRIMEHPC FX10, a
supercomputer capable of world-class
performance of up to 23.2 PFLOPS. The new
system will be available worldwide.
The new system is highly compatible with the
"K computer" supercomputer, which is being
developed by RIKEN and Fujitsu as part of an
initiative led by Japan's Mi nistry of Educatio n,
Culture, Sports, Science and Technology
(MEXT). As such, it is optimal for developing
software to be used on the K computer.
Ultra-high Speed and Ultra-large Scale
Supercomputer
Problems previously constrained or impossible
to solve due to performance limits are now
able to be handled. This is due to the
PRIMEHPC FX10's maximum peak
performance of 23.2 Petaflops and memory
that scales up to 6 PB with a 98,304 node
configuration.
Green Credentials as well as High Performance
Mean Power Savings
In today's quest for a greener world the
compromise between high performance and
environmental footprint is a major issue . A t t he
heart of PRIMEHPC FX10 are SPARC64 IXfx
processo rs that deliver ultra high performance of
236.5 Gigaflops and superb power efficiency of
over 2 Gigaflops per watt.
pplication Performance and Simple
Development
SPARC64 IXfx processo r in clu des extensions for
HPC applications known as HPC-ACE. This plus
wide memory bandwidth, high performance Tofu
interconnect, Technical Computing Suite, HPC
Middl ew are for Petascale Computing, enable
applications to achieve the best performance ever.
In additio n , t he time and effort to adapt to
massively parallel processing is reduced through
the use of VISIMPACT, which simplifies the
implementation of hybrid parallel applications
combining MPI and thread parallelism.
High Reliability and Operability in Large
Systems
Incorporating RAS functions, proven on
mainframe and high-end SPARC64 servers,
SPARC64 IXfx processo r de livers higher
reliability and operability. The flexible 6D
Mesh/Torus architecture of the Tofu
interconnect also contributes to overall
reliability. The result is outstanding
operation: enhanced by the advanced set of
system management, monitoring, and job
management software, and the highly
scalable distributed file system.
*¹ Top-ranked performance: According to the 38
TOP500 list announced on November 2011, with a performance of 10.51 Petaflops.