Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Scribd
Upload
123 views

COSC 4101 Parallel and Distributed Computing Final

The course will cover topics including parallel architectures, interconnection networks, parallel programming models, GPU and CUDA programming, and applications of parallel data processing. Students will complete assignments, projects, tests, and papers on these topics. Evaluation will be based on workshops, quizzes, a semester project, midterm and final papers. Students are expected to attend class regularly and complete all assigned work on time.

Uploaded by

ab2478037
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
123 views

COSC 4101 Parallel and Distributed Computing Final

The course will cover topics including parallel architectures, interconnection networks, parallel programming models, GPU and CUDA programming, and applications of parallel data processing. Students will complete assignments, projects, tests, and papers on these topics. Evaluation will be based on workshops, quizzes, a semester project, midterm and final papers. Students are expected to attend class regularly and complete all assigned work on time.

Uploaded by

ab2478037
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 4

Department of

Computer Science

Tentative Weekly Course Plan

Course Title Parallel and distributed computing


Course Instructor Mr. Saqib Ubaid
Course Code COSC-4101 Credit Hour (3)

Course Objective:

1. With advances in computer architecture, high performance multiprocessor computers have become
readily available & affordable. As a result supercomputing is accessible to a large segment of
industry that was once restricted to military research & large corporations.
2. The course is comprised of architecture & programming of Parallel computing systems; design
concepts, principles, paradigms, models, performance evaluation & applications.

1. Reference Books

i. Parallel Computers: Architecture & Programming by Rajaraman


ii. Introduction to Parallel Computing (2nd ed) by Grama
iii. Programming Massively Parallel Processors (2nd Ed) by Kirk
iv. Programming Massively Parallel Processors (3nd ed)– Hands-on
Approach, by D. Kirk & W. Hwu
v. NVIDIA, NVidia CUDA C Programming Guide, version 4.0, NVidia

TOPICS TO BE COVERED

Week # Topics (outline of main topics and sub topics) Tutorial


/Laboratory

Introduction:
• Introduction to Parallel and Distributed Computing
1
• What, Why and Who uses PDC?
• Visualization and Animation

Interconnection Networks (IN):


• Static INs

2 o Tree
o Hypercube

o k-d Mesh

Parallel Architecture Models:


3 • Implicit:
o Pipelining

Page # 1 / 4
Department of
Computer Science

o Superscalar
o VLIW

Parallel Architecture Models:


• Explicit:
o SIMD
4 o MIMD
o Multicore SMP
o Vector/array processor
o MPP & Clusters
• Parallel Memory Models:
5 o SM, DM, DSM, UMA / NUMA
o Cache Coherence protocols

Interconnection Networks (IN):


• Types of INs
• Static vs Dynamic network topologies
6
• LANs vs SANs (Infiniband)
• Static INs (direct linked topologies)
• Dynamic Ins (Switched networks)
• Heterogeneous (CPU/GPU) Parallel Computing using CudaC

• Multicore SMP programming using OpenMP


7
• Cluster multicomputer programming using MPI

• Group/Parallel Communications & MPI

8 • Combining Parallel Computing (Cluster/GPU) using MPI /


cudaC together

Project Explanation:
9 • Discussion
• Working

Mid Term Exam

• Embedding / Mapping application to IN


10
o Ring/ Mesh into Hypercube & vice versa

• Parallel Data Partitioning Techniques


11
• 1D/ 2D/ 3D; row-wise/ column-wise

Page # 2 / 4
Department of
Computer Science

Graphic Processing Unit (GPU) & cudaC-coding


• Intro to Hetero Computing
12 • Intro to GPU & cudaC
• Data Parallelism Model
•

Graphic Processing Unit (GPU) & cudaC-coding


• CUDA Memory Model
• 1D- Vector addition
• 2D- Image Processing
• 1D & 2D Convolution
13
• Matrix Multiplication
• Square matrices
• Rectangle matrices
• CUDA with MPI

Multidimensional Parallel Data Processing in cudaC


• Image Processing
• Image Convolution
14 • Matrix Multiplication
• revise Image Processing
1D Array Parallel Data Processing Applications
2D-Image Processing Applications

Research:
15 • Presentations

16 Revision

Final Term Exam

Student Evaluation Criteria:


Workshop / Assignments/Case study 2%

Surprise Test/Sudden Test, Quizzes 3%

Semester Project 15%

Mid Term Paper 30%

Final Term paper 50%

Total 100%

Page # 3 / 4
Department of
Computer Science

Student Responsibilities:
Students must attend class. Failure to attend class may result in failure in the course. Students must also arrive on time
and remain in class for the entire period. Cellular Phones and Beeper must be Turned off (Proper classroom
decorum [behavior] adopts, Course outlines and calendars explain requirements and assignments, students are responsible
for knowing what they say. Students are also responsible for doing all assigned work on time. Excessive absences (more
than 03) will result in “F Grade”. Students may prepare Sketchbook for taking notes and for references.

_______________________________
Instructor/ Subject Specialist

____________________________
Chairman/ Head of Department

Page # 4 / 4

You might also like