Information Technology Reference
In-Depth Information
safety of these smaller cars. To address these concerns, automobile engineers use finite element
modeling and massively parallel processing computer systems to simulate crashes. Such sim-
ulations are much less expensive than using actual cars and crash dummies. Also, the speed
and accuracy of the computer simulations allows for many more crash tests at an earlier stage
in the design than using physical crashes with crash dummies. As a result, engineers gain
confidence earlier in the design process that their car will pass federal safety standards, en-
abling them to bring the car to market sooner.
14
Grid computing
is the use of a collection of computers, often owned by multiple indi-
viduals or organizations, to work in a coordinated manner to solve a common problem. Grid
computing is a low-cost approach to parallel computing. The grid can include dozens, hun-
dreds, or even thousands of computers that run collectively to solve extremely large processing
problems. Key to the success of grid computing is a central server that acts as the grid leader
and traffic monitor. This controlling server divides the computing task into subtasks and
assigns the work to computers on the grid that have (at least temporarily) surplus processing
power. The central server also monitors the processing, and if a member of the grid fails to
complete a subtask, it restarts or reassigns the task. When all the subtasks are completed, the
controlling server combines the results and advances to the next task until the whole job is
completed.
European and Asian researchers are using a grid consisting of some 40,000 computers
spread across 45 countries to combat the deadly bird flu. Ulf Dahlsten, a member of the
Information Society and Media Directorate-General of the European Commission, used the
success of grid computing in battling this potential pandemic to point out the breakthroughs
that are being made in drug discovery. “Computer grids have achieved a productivity increase
of more than 6,000 percent in the identification of potential new drugs. Three hundred
thousand molecules have already been screened using the grid. Of these, 123 potential in-
hibitors were identified, of which seven have now been shown to act as inhibitors in in-vitro
laboratory tests. This is a 6 percent success rate compared to typical values of around
0.1 percent using classical drug discovery methods.”
15
Folding@home is a grid computing project with more than 1 million people around the
world downloading and running software to form one of the largest supercomputers in the
world. To carry out their various functions, proteins self-assemble into a particular shape in
a process called “folding.” The goal of the Folding@home project is to research protein
folding and misfolding and gain an understanding of how this protein behavior is related to
diseases such as Alzheimer's, Parkinson's, and many forms of cancer. It takes a single com-
puter about one day to simulate a nanosecond (1/1,000,000
th
of a second) in the life of a
protein. The folding process takes about 10,000 nanoseconds. Thus, 10,000 days (30 years)
are required to simulate a single folding! The Folding@home group has developed ways
to speed up the simulation of protein folding by dividing the work among over 100,000
ticipating on the grid, the combined computing power exceeded 1 × 10
15
floating-point
operations per second—more than twice the speed of the world's fastest stand-alone super-
computer.
17
The most frequent uses for parallel computing include modeling, simulation, and
analyzing large amounts of data. Chrysler uses high-performance computers consisting of
some 1,650 cores to simulate racecar performance and identify opportunities for improve-
ment in the car's design and operation. The ability to develop more complete fluid dynamic
models of the extreme conditions associated with vehicles traveling at 190 mph has led to
improvements not only in racecars but in passenger cars as well. For example, simulations
show how a racecar traveling behind another car receives restricted airflow, which can affect
engine performance. This finding can be reapplied to the design of passenger cars to deal
with the restricted airflow they receive when traveling behind a large truck.
18
Cloud computing
involves using a giant cluster of computers that serves as a host to run
applications that require high-performance computing. Cloud computing supports a wider
variety of applications than grid computing and pools computing resources so they can be
managed primarily by software rather than people. IBM and Google have provided hardware,
software, and services to many universities so that students and faculty can explore cloud
grid computing
The use of a collection of computers,
often owned by multiple individuals
or organizations, to work in a coor-
dinated manner to solve a common
problem.
cloud computing
Using a giant cluster of computers
to serve as a host to run applications
that require high-performance
computing.
