CS481 Parallel and

Distributed
Computing
Instructor
Muhammad Danish Khan
Lecturer, Department of Computer Science
FAST NUCES Karachi
m.danish@nu.edu.pk

Recommended Text Book

Introduction to Parallel Computing, Second Edition

by Ananth Grama
Marks Distribution and
Deadlines
Assessment Weightage Schedule/Deadlines/ Remarks
Mid-term I Examinations 15 6th Week
Mid-term II Examinations 15 12th Week
Tasks/Assignments 10 1-2 Assignment Tasks per Week with Plagiarism
Checking
Semester Project 10 Project idea submission: 7th week, Final Submission
due in 15th - 16th Week
Final Examinations 50 Comprehensively from all the covered and assigned
topics.
Pre-Requisite
Operating System Concepts
Algorithms
Semester Plan
Week 1: Revision of Operating System Concepts

Week 2: Introduction to Parallel Computing, Assignment Task(s)

Week 3: Parallel Programming Platforms

Week 4: Parallel Programming Platforms , Assignment Task(s)

Week 5: Principles of Parallel Algorithm Design, Quiz-1

Week 6: Mid Term-1 Examinations

Semester Plan
Week 7: Principles of Parallel Algorithm Design, Assignment Task(s), Project Proposal Submission

Week 8: Programming Shared Address Space

Week 9: Programming Shared Address Space

Week 10: Programming Shared Address Space, Assignment Task(s)

Week 11: Programming Using the Message Passing Paradigm, Quiz-2

Week 12: Mid Term-2 Examinations

Semester Plan
Week 13: Programming Using the Message Passing Paradigm, Assignment Task(s)

Week 14: Distributed System Models and Enabling Technologies, Assignment Task(s)

Week 15: Distributed System Models and Enabling Technologies, Quiz-3, Project
Evaluations

Week 16: Distributed System Models and Enabling technologies, Project Evaluations
LMS: Google Class Room
Section 5C Class Code: eg2u47t
https://classroom.google.com/c/Mzg4NTA1MTMzNDI3?cjc=eg2u47t
Operating System Concepts
Program
◦ Set of instructions and associated data
◦ resides on the disk and is loaded by the operating system to perform some task.
◦ E.g. An executable file or a python script file.

Process
◦ A program in execution.
◦ In order to run a program, the operating system's kernel is first asked to create a new
process, which is an environment in which a program executes.
◦ consists of instructions, user-data, and system-data segments, CPU, memory, address-
space, disk acquired at runtime
Thread
◦ the smallest unit of execution in a process.
◦ A thread simply executes instructions serially.
◦ A process can have multiple threads running as part of it.
◦ Processes don't share any resources amongst themselves whereas threads
of a process can share the resources allocated to that particular process,
including memory address space.
CLOUD COMPUTING DEFINITION

•Paralllel computing (processing):

• the use of two or more processors (computers), usually
within a single system, working simultaneously to solve a
single problem.

•Distributed computing (processing):

• Any computing that involves multiple computers remote
from each other that each have a role in a computation
problem or information processing.
Parallel programming:
• The human process of developing programs that express what
computations should be executed in parallel.
Serial Computing:

•Traditionally, software has been written for serial

computation:
•A problem is broken into a discrete series of instructions
•Instructions are executed sequentially one after another
•Executed on a single processor
•Only one instruction may execute at any moment in time
Parallel Computing
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.
Serial execution
“Multiprocessing systems”
◦ where multiple processes get scheduled on more than one CPU.
◦ Usually, this requires hardware support where a single system comes with
multiple cores
◦ or the execution takes place in a cluster of machines.

Multiple Processors vs Multiple Cores

Concurrency vs Parallel
Execution
Concurrency

Parallel Execution
Parallelism
The term parallelism means that an application splits its tasks up
into smaller subtasks which can be processed in parallel, for instance
on multiple CPUs at the exact same time.
 Serial Execution vs. Parallel Execution

Concurrent Execution on a Single-core System

Parallel Execution on a Multicore System

 Limitations of Serial
Computing
Limits to serial computing - both physical and practical reasons pose significant
constraints to simply building ever faster serial computers.

Transmission speeds - the speed of a serial computer is directly dependent upon how
fast data can move through hardware.
◦ Absolute limits are the speed of light (30 cm/nanosecond) and the transmission limit of copper
wire (9 cm/nanosecond). Increasing speeds necessitate increasing proximity of processing
elements.

Economic limitations - it is increasingly expensive to make a single processor faster.

◦ Using a larger number of moderately fast commodity processors to achieve the same (or better)
performance is less expensive.
Parallel Computing
In the simplest sense, parallel computing is the
simultaneous use of multiple computing resources
to solve a computational problem.
◦ To be run using multiple CPUs
◦ A problem is broken into discrete parts that can be solved
concurrently
◦ Each part is further broken down to a series of instructions

Instructions from each part execute simultaneously

on different CPUs
Parallel Computing: Resources
The compute resources can include:
◦ A single computer with multiple processors;
◦ A single computer with (multiple) processor(s) and some specialized
computer resources (GPU, FPGA …)
◦ An arbitrary number of computers connected by a network (Cluster)
◦ A combination of both.
Parallel Computing: The
computational problem
The computational problem usually demonstrates characteristics
such as the ability to be:
◦ Broken apart into discrete pieces of work that can be solved simultaneously;

◦ Execute multiple program instructions at any moment in time;

◦ Solved in less time with multiple compute resources than with a single
compute resource.
LD $12, (100)
ADD $11, $12
SUB $10, $11
INC $10
SW $13, ($10)
int sample1
{
X = sample2()
Return x;
}
float sample3
{
Pi=3.14
Return pi
}
Int sample2()
{
Cin>>I
Return I;
}
Parallel Computing: what for?
 Example applications include:

◦ Parallel Databases, Data Mining

◦ Web Search Engines, Web Based Business Services

◦ Computer-aided diagnosis in medicine

◦ advanced graphics and virtual reality, particularly in the entertainment industry

◦ networked video and multi-media technologies

Why Parallel Computing?
Save time

Solve larger problems

Provide parallelism (do multiple things at the same time)

… ..
Flynn Taxonomy
referred to as Flynn's Taxonomy, is a classification system for
computer architectures introduced by Michael J. Flynn in 1966. It
categorizes computer systems based on the number of instruction
streams and data streams they can handle simultaneously. This
taxonomy is particularly useful for understanding parallel processing
and the design of processors.
Flynn Taxonomy
Based on the number of concurrent instruction (single or multiple)
and data streams (single or multiple) available in the architecture
Single Instruction, Single Data
(SISD)
It represents the organization of a single computer containing a control
unit, processor unit and a memory unit.

Single instruction: only one instruction stream is being acted on by the

CPU during any one clock cycle
Single data: only one data stream is being used as input during any one
clock cycle

A single processor executes a single instruction stream, to operate

on data stored in a single memory.

This is the oldest and until recently, the most prevalent form of computer
◦ Examples: most PCs, single CPU workstations and mainframes
Single Instruction, Single Data
(SISD)
Single Instruction, Multiple Data
(SIMD)
Single instruction: All processing units execute the same instruction at any given clock cycle
Multiple data: Each processing unit can operate on a different data element
The processing units are made to operate under the control of a common control unit, thus
providing a single instruction stream and multiple data streams.
◦ Best suited for specialized problems characterized by a high degree of regularity, such as image processing.

Two varieties: Processor Arrays and Vector Pipelines

Examples:
◦ Processor Arrays: Connection Machine CM-2, Maspar MP-1, MP-2
◦ Vector Pipelines: IBM 9000, Cray C90, Fujitsu VP, NEC SX-2, Hitachi S820
Single Instruction, Multiple Data
(SIMD)
 Multiple Instruction, Single Data
(MISD)
A single data stream is fed into multiple processing units.

It consists of a single computer containing multiple processors connected with

multiple control units and a common memory unit

Each processing unit operates on the data independently via

independent instruction streams.

Some conceivable uses might be:

◦ multiple frequency filters operating on a single signal stream
Multiple Instruction, Single Data
(MISD)
Multiple Instruction, Multiple
Data (MIMD)
Currently, the most common type of parallel computer. Most
modern computers fall into this category.
It represents the organization which is capable of processing
several programs at same time.
Multiple Instruction: every processor may be executing a
different instruction stream
Multiple Data: every processor may be working with a
different data stream
Execution can be synchronous or asynchronous,
deterministic or non-deterministic
◦ Examples: most current supercomputers, networked parallel computer
"grids" and multi-processor SMP computers - including some types of PCs.
Multiple Instruction, Multiple
Data (MIMD)
Some General Parallel
Terminology
Task/Process
◦ A logically discrete section of computational work. A task is typically a program or program-
like set of instructions that is executed by a processor.

Parallel Task
◦ A task that can be executed by multiple processors safely (yields correct results)

Serial Execution
◦ Execution of a program sequentially, one statement at a time. In the simplest sense, this is
what happens on a one processor machine. However, virtually all parallel tasks will have
sections of a parallel program that must be executed serially.
Parallel Execution
◦ Execution of a program by more than one task, with each task being able to execute the same or
different statement at the same moment in time.
Shared Memory
◦ From a strictly hardware point of view, describes a computer architecture where all processors have
direct (usually bus based) access to common physical memory.
◦ In a programming sense, it describes a model where parallel tasks all have the same "picture" of memory and can directly address and access the same
logical memory locations regardless of where the physical memory actually exists.

Distributed Memory
◦ In hardware, refers to network based memory access for physical memory that is not common. As a
programming model, tasks can only logically "see" local machine memory and must use
communications to access memory on other machines where other tasks are executing.
Communications
◦ Parallel tasks typically need to exchange data. There are several ways this can be accomplished, such as
through a shared memory bus or over a network, however the actual event of data exchange is
commonly referred to as communications regardless of the method employed.

Synchronization
◦ The coordination of parallel tasks in real time, very often associated with communications. Often
implemented by establishing a synchronization point within an application where a task may not
proceed further until another task(s) reaches the same or logically equivalent point.

◦ Synchronization usually involves waiting by at least one task, and can therefore cause a parallel
application's wall clock execution time to increase.
Granularity
◦ In parallel computing, granularity is a qualitative measure of the ratio of computation to communication.
◦ Coarse: relatively large amounts of computational work are done between communication events
◦ Fine: relatively small amounts of computational work are done between communication events

Observed Speedup
◦ Observed speedup of a code which has been parallelized, defined as:

◦ One of the simplest and most widely used indicators for a parallel program's performance.
Parallel Overhead
◦ The amount of time required to coordinate parallel tasks, as opposed to doing useful work. Parallel
overhead can include factors such as:
◦ Task start-up time
◦ Synchronizations
◦ Data communications
◦ Software overhead imposed by parallel compilers, libraries, tools, operating system, etc.
◦ Task termination time

Massively Parallel
◦ Refers to the hardware that comprises a given parallel system - having many processors. The meaning of
many keeps increasing, but currently BG/L* pushes this number to 6 digits.

*Blue Gene is an IBM project aimed at designing supercomputers that can reach operating
speeds in the petaFLOPS (PFLOPS) range, with low power consumption.
Scalability
◦ Refers to a parallel system's (hardware and/or software) ability to
demonstrate a proportionate increase in parallel speedup with the addition of
more processors.

◦ Factors that contribute to scalabilty include:

◦ Hardware - particularly memory-cpu bandwidths and network communications
◦ Application Algorithm
◦ Parallel overhead related
◦ Characteristics of your specific application and coding
Parallel Computer
Memory
Architectures
Memory architectures
Shared Memory
Distributed Memory
Hybrid Distributed-Shared Memory
 Shared Memory
Shared memory parallel computers vary widely, but generally have in common the ability for all processors to access all
memory as global address space.

Multiple processors can operate independently but share the same memory resources.
Changes in a memory location effected by one processor are visible to all other processors.
Shared memory machines can be divided into two main classes based upon memory access times: UMA and NUMA.
Shared Memory : UMA vs. NUMA
Uniform Memory Access (UMA):
◦ Most commonly represented today by Symmetric Multiprocessor (SMP) machines
◦ Identical processors with equal access and access times to memory
◦ Sometimes called CC-UMA - Cache Coherent UMA.

Non-Uniform Memory Access (NUMA):

◦ Often made by physically linking two or more SMPs
◦ One SMP can directly access memory of another SMP
◦ Not all processors have equal access time to all memories
Shared Memory: Pro and Con
Advantages
◦ Global address space provides a user-friendly programming perspective to memory
◦ Data sharing between tasks is both fast and uniform due to the proximity of memory to CPUs
Disadvantages:
◦ Primary disadvantage is the lack of scalability between memory and CPUs. Adding more CPUs can
geometrically increases traffic on the shared memory-CPU path, and for cache coherent systems,
geometrically increase traffic associated with cache/memory management.
◦ Programmer responsibility for synchronization constructs that insure "correct" access of global memory.
◦ Expense: it becomes increasingly difficult and expensive to design and produce shared memory
machines with ever increasing numbers of processors.
 Distributed Memory
Like shared memory systems, distributed memory systems vary widely but share a common characteristic. Distributed
memory systems require a communication network to connect inter-processor memory.
Processors have their own local memory. Memory addresses in one processor do not map to another processor, so
there is no concept of global address space across all processors.
Because each processor has its own local memory, it operates independently. Changes it makes to its local memory have
no effect on the memory of other processors. Hence, the concept of cache coherency does not apply.
When a processor needs access to data in another processor, it is usually the task of the programmer to explicitly define
how and when data is communicated. Synchronization between tasks is likewise the programmer's responsibility.
The network "fabric" used for data transfer varies widely, though it can can be as simple as Ethernet.
Distributed Memory: Pro and
Con
Advantages
◦ Memory is scalable with number of processors. Increase the number of processors and the size of
memory increases proportionately.
◦ Each processor can rapidly access its own memory without interference and without the overhead
incurred with trying to maintain cache coherency.
◦ Cost effectiveness: can use commodity, off-the-shelf processors and networking.

Disadvantages
◦ The programmer is responsible for many of the details associated with data communication between
processors.
◦ It may be difficult to map existing data structures, based on global memory, to this memory
organization.
◦ Non-uniform memory access (NUMA) times
Hybrid Distributed-Shared
Memory
Comparison of Shared and Distributed Memory Architectures

Architecture CC-UMA CC-NUMA Distributed

Examples SMPs Bull NovaScale Cray T3E

Sun Vexx SGI Origin Maspar
DEC/Compaq Sequent IBM SP2
SGI Challenge HP Exemplar IBM BlueGene
IBM POWER3 DEC/Compaq
IBM POWER4 (MCM)

Communications MPI MPI MPI

Threads Threads
OpenMP OpenMP
shmem shmem

Scalability to 10s of processors to 100s of processors to 1000s of processors

Draw Backs Memory-CPU bandwidth Memory-CPU bandwidth System administration

Non-uniform access times Programming is hard to develop
and maintain
Software Availability many 1000s ISVs many 1000s ISVs 100s ISVs
Assignment:
For this assignment, you and your partner will research a specific
parallel or distributed system and create a detailed description of
your system. Additionally, you will prepare a 12-minute presentation
to present to the class. You will need to prepare 5-7 slides for your
presentation, ensuring that your talk lasts no more than 10 minutes,
leaving 2 minutes for questions.
For your system, you should address each of the items listed below.

1. .Definition of the system

2. Architecture overview. For this part you can grab figures from other
sites (just be sure to list a citation of from where you grabbed a figure if
you do this)
3. How are processors/nodes interconnected?
4. What things are shared? what are private?
5. Is there a specific programming model for this system? explain. Are
there some programming models that are not supported by this
system? explain.
6. Is the system designed/optimized for a specific type of use? or is it
designed for a specific type of program workload? explain.
7. How scalable is the system? explain.