Chapter 06: Advanced Topics

Hari K.C.
Pashchimanchal Campus, IOE, TU
Multiprocessing Systems

• Multiprocessing is an example of parallelism which uses multiple

CPU’s sharing the common resources such as memory, storage device
etc.
• As computer technology has evolved and as the cost of computer
hardware has dropped, computer designers have sought more and
more opportunities for parallelism, usually to enhance performance
and availability.
• Instruction pipelining and the overlapping of fetch & execute
instructions from the same program in parallel.
Characteristics of Multiprocessing system

• Contains two or more similar general purpose processors of

comparable capability.
• All processors share access to common memory.
• All processors share access to I/o devices either through the same
channels that provide paths to the same devices.
• System is controlled by an integrated operating system that provides
interaction between processors and their programs.
Here the processors can communicate with each other through memory. The CPUs can
directly exchange signals as indicated by dotted line.
The organization of multiprocessor system can be divided into three types.

1. Time shared or common bus:

• There are number of CPUs, I/O modules and memory modules connected to the same bus.
• So the time shared system must distinguish the modules on the bus to determine the source and
destination of the data.
• Any module in the bus can temporarily act as a master. When one module is controlling the bus,
the other should be locked out.
• The access to each module is divided on the basis of time. The time shared multiprocessing
system has the following advantages.
• Simplicity:
The physical interface and the addressing time sharing logic of each processor remains the same as
in a single processor system, so it is very simplest approach.
• Flexibility:
It is easy to expand the system by attaching more CPUs to the bus.
• Reliability:
The failure of any attached device should not the failure of the whole system.
Drawback:
The speed of the system is limited by the cycle time because all memory references must pass
through the common bus.
2. Multiport memory:
• Each processor and /O module has dedicated path to each memory
module this system has more performance and complexity than
earlier one.
• For this system, it is possible to configure portions of memory as
private to one or more CPUs and or I/O modules.
• This feature allows increasing security against unauthorized access
and the storage of recovery routines in areas of memory not
susceptible to modification by other processors.
3. Central control unit :

• It manages the transfer of separate data streams back and forth

between independent modules like CPU, memory and I/o.
• the controller can buffer requests and perform arbitration and timing
functions.
• It can also pass status and control messages between CPU’s. All the
co-ordination is concentrated in the central control unit un-disturbing
the modules.
• It is more flexible and complex as well.
Real and Pseudo-Parallelism
• Traditionally, software has been written for serial computation:
- To be run on a single computer having a single Central Processing Unit
(CPU).
- A problem is broken into a discrete series of instructions.
- Instructions are executed one after another.
- Only one instruction may execute at any moment in time.
• In the simplest sense, parallelism is the simultaneous use of multiple
compute resources to solve a computational problem:
- To be run using multiple CPU
- A problem is broken into discrete parts that can be solved concurrently o
- Each part is further broken down to a series of instructions
- Instructions from each part execute simultaneously on different CPUs
• Real parallelism consists of the parallel modes of physical devices so
that each can carry parallel operations to each other.
• Core parallelism consists of real parallelism. Multiple core processes
which are physically different and performs their own operations in
parallel.
• Pseudo parallelism consists of the same device carrying the parallel
operation.
• We can logically manage the parallelism for system. Concurrent
processing using parallelism is the pseudo parallelism which operates
either in time division or using other types of parallel algorithms.
Flynn Classification

• There are different ways to classify parallel computers. One of the

more widely used classifications, in use since 1966, is called Flynn's
classification.
• Flynn's classification distinguishes multi-processor computer
architectures according to how they can be classified along the two
independent dimensions of Instruction and Data.
• Each of these dimensions can have only one of two possible states:
Single or Multiple.
• The matrix below defines the 4 possible classifications according to
Flynn:
Instruction Level, Thread Level and Process Level
Parallelism
Inter-process Communication, Resource Allocation
and Deadlock
Features of Typical Operating System
Different Microprocessor Architectures
RISC and CISC
Digital signal processor
• The real time signals such as pressure, temperature, voice are
continuous time varying signals are known as analog signals.
• The process of conversion of the analog signals into discrete signal
means digital signal which reduces the redundancy and make more
immune to noise is called digital signal processing .the digital signal
processors take input the digital data for that every analog data is to
be converted into digital by using A\D converter.
• For example microphone converts sound into electrical energy i.e.
analog, it is sampled by A\D converter and converted into digital one
which is fed to the DSP processor.
• After performing these data DSP can transfer the discrete data to D\A
converter which further to speaker to convert electrical signal to sound.
• The whole function is carried out by DSP processers using hardware like
microphone, transducer, A\D converter, D\A converter, speaker etc. and
software like C or MATLAB which carried out FFT (fast Fourier transform).
DSP processors have low processing speed due to very curtail signals need
to be operated.
• The micro-processers and computers we used today are based on non
Neumann architecture where the instruction defines both the operation
and data.
• So the DSP processors should be fast processing and for that we need to
design best architecture which follows Harvard Architecture where
separate buses for instructions and data are used.
• DSP chips are specially designed for particular application and they are not
used for general type of processing like microprocessors do.
• There are very few manufacturers for DSP chip, one of them is the Texas
instruments, USA. Its TMS320C series is worldwide popular and can be
used for implementing various types of signal processing application.
Applications of a Digital signal processing system
• Telecommunication
• For echo cancellation.
• Equalization – Think about tuning your radio for bass and treble).
• Filtering – Removing unwanted signals
• Instrumentation and Control
• In designing Phase Locked Logic (PLL).
• Noise reduction circuits.
• Compression of signals.
• Function generators.
• Digital Image Processing
• Compression of an image.
• Enhancement, reconstruction, and restoration of an image.
• Speech Processing
• Digital audio synthesis.
• Speech recognition and analysis.
• Medicine
• X-rays, ECGs, EEGs.
• Signal filtering
• Noise removal and shaping of signal spectrums.
The end