By SAREN JOHNSTON
Ames Lab Public Affairs
Bunch them, group them, gather them together -- personal computer clusters
are making parallel power far less costly than traditional supercomputing systems.
Supercomputers are just as powerful as their name inplies. They contain
several to many thousand central processing units (CPUs) and come in large boxed
systems. They have impressive computing power and heart-stopping prices. Just
one such system can cost millions of dollars. And don't forget to figure in 10 to 20
percent of the purchase price for yearly maintenance.
But high-performance parallel computing doesn't have to cost a fortune, says
Mark Gordon, Ames Laboratory program director for Applied Mathematics and
Computational Sciences. He and several of his colleagues at Ames Laboratory's
Scalable Computing Lab (SCL) are networking personal computers (PCs) in clusters of
eight, 16, 32 and 64 to create parallel computing systems comparable to today's best
supercomputers, and for a fraction of the cost.
Parallel computers are like valued employees, they can handle a number of
tasks at the same time. Their multiple processors work simultaneously on different
parts of a single problem, making them far more efficient and able to handle more
complex problems than sequential computers, which tackle a problem one step at a
time.
Gordon and fellow SCL researchers David Halstead, John Gustafson,
Stephen Elbert, Don Heller, Dave Turner and Bruce Harmon are hoping their research
on PC clusters will make parallel computing more economical and attainable for
scientific and educational communities. They are devising a "cluster cookbook"
for
the world wide web, which will provide guidelines on how to construct PC clusters.
They will also develop and facilitate a hands-on workshop with the goal of bringing
the cost-saving cluster computing technique into university departments, individual
research groups and the classroom.
The clustering effort came about as PC manufacturers tried to outdo one
another by incorporating more and more supercomputing ideas into their desktop
machines. "In the process, they made these little computers competitive with
workstations," says Gordon. "But what's hard is networking clusters of these
sequential computers to make a true, parallel, high-performance computer that's
competitive with boxed systems that cost millions of dollars."
The SCL researchers have made that intricate job look easy. "It's clear that
small clusters work," says Gordon. "We have a math cluster with eight nodes
running
partial differential equations. And all of our quantum chemistry codes can run in
parallel." Gordon adds that in other areas of the Lab, clusters of eight PCs are
doing
materials simulations and modeling new materials with desirable magnetic properties.
"So the issue is does clustering work when you scale up to 64; does it work when you
scale up to 128. That's really unknown."
Whether larger networks of PC clusters will work depends to a great extent
on how well researchers can optimize message-passing among the various computers
in a cluster. It's not difficult to imagine that the arrangement of PCs affects how well
communication takes place. And the more arrangements you have, the more traffic-
directing problems you get.
Halstead reminds us of a more visible issue. "There's a definite storage
concern for a large cluster network. Where do you put the thing? Also, all the energy
that goes into a computer comes out as heat at the end of the day, so you'll use more
power for air conditioning."
SCL researchers believe the potential benefits of scalable cluster computing
far outshine the issues they are working to resolve. A big advantage of taking the
cluster path to parallel computing is that you can always reconfigure the system to
meet the needs of the day. "You can set these things up as a computer lab, launch
them to run a word processing program for a class, and reboot them to run as a single
parallel computer. You can trade off with other departments -- whatever you want to
do," says Halstead. "With this kind of sharing of resources, you're only really
limited
by curriculum originality."
Halstead also notes that although PC clusters are not meant to replace the
powerful supercomputers that operate at Department of Energy national labs, they can
lessen the computing burden placed upon these machines and so reduce the barrier to
national computer use. Without a doubt, however, low cost is the biggest advantage.
"Cluster computing is a happy synchronization of technologies," says
Halstead. "If a PC dies, you throw it away and go down and buy another one, kind of
like replacing a fan belt. The beauty is if a particular PC is not needed in a cluster,
it's
still a very powerful desktop machine. You're not buying into a hardware technology
that in a year's time will be completely useless due to lack of software support."
Maybe the answer to whether clusters will work when scaled up to larger and
larger systems will come soon. The SCL team recently constructed a cluster of 64 PCs,
each with two central processing units, and are now testing its ability to perform
parallel computations.
"At the end of the day, this thing should be four times as fast and have four
times the storage capacity and memory as the largest supercomputer in the SCL,
which cost just shy of a million dollars," says Halstead. "So it's four times as
fast for a
third of the price."
The SCL researchers have compared the performance of their clusters with
that of commercial parallel computers by using a computer benchmark called HINT,
which Gustafson developed and for which he earned an R&D 100 Award in 1995. "I
think HINT is better than just about any other way of benchmarking computers," says
Gordon. "If you take John's method and evaluate our clusters against one of the best
parallel computers on the market today and then go the next step and divide that by
the cost to get the price/performance ratio, the clusters blow everything else away.
And that's our point -- not to show people how to do really great parallel computing
because lots of people can do that, but to show them how they can do it in a very
cost-effective way."
Related material:
Ames Lab at forefront of cluster computing
Last revision: 6/2/98 sd
Home | Comments | Search | Disclaimer