3. 1.1 Fundamentals
• Parallel Computers can be classified according to
their machine structure.
• Parallel operations may be grouped by their
abstraction level or by the types of argument.
– 1. Sequential programs is special case of
parallel programming.
4. – 2. Petri nets is useful tool for definition
asynchronous parallelism.
– 3. Petri nets and other (flow diagrams are
used to discover data dependencies before
designing an parallel program.
– 4. The most common parallel programming
languages and connection structures in a
parallel computer system are fundamental for
understanding parallel computations.
• The growth of a plant is simultaneously influenced
by number of factors.
• The world is parallel; Natural systems , complex
technical processes and even societal chances are
highly parallel processes.
5. – 1. Quantum chemistry, stoical mechanics,
relativistic physics.
– 2. Cosmology and astrophysics.
– 3. Computational fluid dynamics and
turbulence.
– 4. Material design an super conductivity.
– 5. Biology, genome sequencing, genetic
engineering, protein folding, enzyme activity,
and cell modeling.
– 6. Medicine and modeling human organs and
bones.
– 7. Global weather and environmental
modeling.
6. – 8. Photo-realistic computer graphics
Geological time), pattern and etc. recognition,
(Parallel and Distributed Computing is
necessary).
• For example; (1994) a single iteration of ocean
circulation takes 10 minutes and 30 billion
floating point calculation. (50 million 3-D cell).
• Molecular dynamics of a protein and surrounding
water would take 31,688 years on Cray X/MP.
• Parallel computers, Supercomputers,
• Parallel processing, Multi-threading, Throughput,
Pipelined (segments and stages) computations.
8. • However, many parallel super computer are
developed.
• Simple personal computer may carry out may
operation simultaneously, multiple I/O via DMA
and simple CPU has many parallel functional units
Coprocessors.
• Even arithmetic operation can be divided into
parallel operations
• Figure 1.1 Explains Von Neumann and Human
Brain (Page 3)
• Table 1.2 Illustrates incongruity of Von Neumann
computer.
1.3 Parallel Computers
9.
10.
11. – Parallel computing is natural form of
information processing.
– Sequential Von Neumann computing successful
but it is an artificial restriction.
– Parallel representation of a problem has more
information from the problem itself. Sequential
can easily be constructed.
– The goal of a parallel programming
understandable programs and fast execution
times.
