IBM, Lab to work on new design for Blue Gene
IBM last week announced a partnership with the Department of Energy’s
National Nuclear Security Agency to expand IBM’s Blue Gene research
project.
IBM will work with LLNL to jointly design a new supercomputer in the Blue
Gene family. Called Blue Gene/L, the machine will be at least 15 times
faster, 15 times more power efficient and consume about 50 times less
space per computation than today’s fastest supercomputers.
Blue Gene/L marks a major expansion of the Blue Gene project. Blue Gene/L
is expected to operate at about 200 teraflops (200 trillion operations
per second) which is larger than the total computing power of the top
500 supercomputers in the world today. Blue Gene/L will also be a part
of IBM’s research in "autonomic computing,’’ an initiative
to design computer systems that are self-healing, self-managing and self-configuring.
The development of Blue Gene/L will take place as part of work under way
to build a petaflop-scale (one quadrillion operations per second) machine,
as announced in December 1999.
IBM and the Laboratory have a long history of working together on supercomputing
projects, most significantly on the Accelerated Strategic Computing Initiative
(ASCI) Program. IBM built the "ASCI White’’ machine for
the Laboratory, which is the world’s current record-breaking supercomputer.
"Our initial exploration made us realize we can expand our Blue Gene
project to deliver more commercially viable architectures for a broad
customer set, and still accomplish our original goal of protein science
simulations,’’ said Mark Dean, vice president of systems, IBM
Research. "Partnering with Lawrence Livermore is a key part of our
strategy, as they bring important application and design expertise to
the project.’’
Researchers at the national laboratories plan to use Blue Gene/L, which
is expected to be completed by 2005, to simulate physical phenomena of
national interest — such as aging of materials, fires, and explosions
— that require computational capability much greater than presently
available.
"This represents a new thrust, very different from the approach taken
by the main line of ASCI machines. Up until now, ASCI supercomputers have
been designed to address the entire spectrum of numerical simulations
required of the stockpile stewardship effort,’’ said David Nowak,
ASCI Program Leader at LLNL. "This new Blue Gene/L innovation can
address an important subset of those computational problems, those that
can be easily divided to run on many tens of thousands of processors.’’
"Examples of those applications include the modeling of the aging
and properties of materials, and the modeling of turbulence,’’
added Nowak. "This technology opens the door to a number of applications
of great interest to civilian industry and business, like biology and
other life sciences. The future of U.S. high-performance computing will
benefit tremendously from pursuing both of these paths in parallel.’’
New architecture for supercomputers
While today’s machines are amazingly fast number crunchers, many
data-intensive applications are slowed because of the time it takes to
simply access information from the memory chips. The Blue Gene/L design
will run these applications much faster because the machine will be populated
with data-chip cells optimized for data access. Each chip includes two
processors: one for computing and one for communicating, and its own on-board
memory. Each of the data-chip cells will work on a small part of a larger
problem. This increase in data access speed will make a huge difference
in the kinds of results these machines can produce and the kinds of problems
they can solve.
NNSA’s Bill Reed, ASCI’s national program leader, lists an impressive
array of projects that can make use of this new approach and cites "the
continuing need for cost-effective computing to address important national
security issues. We need to run these problems in days not months and
we need to simultaneously support many scientists across all three NNSA
laboratories working on a broad spectrum of technical issues. The value
to both national security programs and commercial interests can be dramatic,
especially in the biological sciences and medical and pharmaceutical fields.’’
IBM and the Laboratory will team up to explore the hardware and software
components needed to construct this new computing architecture, and Livermore
will provide additional design expertise for the applications that can
take advantage of the Blue Gene/L machine.
Lawrence Livermore will get help on the Blue Gene/L project from collaborators
at the DOE’s NNSA, Columbia University, San Diego Supercomputing
Center, and Caltech.
The ASCI Program at Lawrence Livermore, Los Alamos and Sandia national
laboratories has been partnering with the supercomputing industry for
the past five years in developing a series of supercomputers for NNSA’s
Stockpile Stewardship Program. This latest effort continues to build on
that experience to help enable the United States to maintain its nuclear
stockpile without underground nuclear testing and make unprecedented contributions
to many fields of science that rely heavily on computing and simulation.