Hardware: Input, Processing, and Output Devices - Principles of Information Systems - page 95

Multiprocessing
Generally, multiprocessing involves the simultaneous execution of two or more instructions
at the same time. One form of multiprocessing uses coprocessors. A coprocessor speeds pro-
cessing by executing specific types of instructions while the CPU works on another processing
activity. Coprocessors can be internal or external to the CPU and can have different clock
speeds than the CPU. Each type of coprocessor performs a specific function. For example, a
math coprocessor chip speeds mathematical calculations, and a graphics coprocessor chip
decreases the time it takes to manipulate graphics.
A multicore microprocessor combines two or more independent processors into a single
computer so that they can share the workload and boost processing capacity. In addition, a
dual-core processor enables people to perform multiple tasks simultaneously, such as playing
a game and burning a CD. The use of low clock speed, multicore processors with a shared
cache (rather than separate dedicated caches for each processor core) is another approach to
reduce the heat generated by the computer without reducing its processing power. For
example, the Intel Dual Core processor runs at 1.66 MHz compared to the single core
Pentium 4 processor, which runs at 3.2 GHz. AMD and Intel are battling for leadership in
the multicore processor marketplace.
Both Intel and AMD have improved on dual processors by introducing new quad-core
chips. However, a major need for basic research in computer science is to develop software
that can actually take advantage of four processors. The processor manufacturers are working
with software developers to create new multithreaded applications and next-generation
games that will use the capabilities of the quad-core processor. “Industry has basically thrown
a Hail Mary,” warns David Patterson, a pioneering computer scientist at the University of
California, Berkeley, referring to the hardware shift. “The whole industry is betting on parallel
computing. They've thrown it, but the big problem is catching it.” 10
Multicore systems are most effective when they run programs that can split their work-
load among multiple CPUs. Such applications include working with large databases and
multimedia. Intel has introduced Viiv (rhymes with five), a marketing initiative that com-
bines Intel products including the Core 2 Quad processor with additional hardware and
software to build an extremely powerful multimedia computer capable of running the highly
processing-intensive applications associated with high-definition entertainment, including
the following: 11
• Full 1080P video playback of movie clips, media streams, and HD video cameras
• High-quality audio for surround-sound capabilities from movies and music
• Fast, extremely high-quality photo editing, retouching, and publishing
• Capability to watch, record, and pause live TV
AMD has countered with its new quad-core Opteron processor. AMD also plans to catch
up to Intel by building its first 45-nanometer chips in 2008. 12 AMD is counting on a project
it calls Fusion, which will combine a graphics processing unit and a CPU on the same chip,
and is expected in late 2008 . 13
When selecting a CPU, organizations must balance the benefits of processing speed with
energy requirements and cost. CPUs with faster clock speeds and shorter machine cycle times
require more energy to dissipate the heat generated by the CPU and are bulkier and more
expensive than slower ones.

multiprocessing
The simultaneous execution of two
or more instructions at the same
time.

coprocessor
The part of the computer that
speeds processing by executing
specific types of instructions while
the CPU works on another process-
ing activity.

multicore microprocessor
A microprocessor that combines
two or more independent proces-
sors into a single computer so they
can share the workload and improve
processing capacity.

parallel computing
The simultaneous execution of the
same task on multiple processors to
obtain results faster.

Parallel Computing
Parallel computing is the simultaneous execution of the same task on multiple processors
to obtain results faster. Systems with thousands of such processors are known as massively
parallel processing systems . The processors might communicate with one another to coor-
dinate when executing a computer program, or they might run independently of one another
but under the direction of another processor that distributes the work to the other processors
and collects their processing results. The dual-core processors mentioned earlier are a simple
form of parallel computing.
In response to higher fuel prices and the desire to reduce carbon dioxide emissions, auto
manufacturers are introducing smaller models. However, consumers are concerned about the

massively parallel processing
systems
A form of multiprocessing that
speeds processing by linking hun-
dreds or thousands of processors to
operate at the same time, or in par-
allel, with each processor having its
own bus, memory, disks, copy of the
operating system, and applications.