Scribd
Upload
75 views11 pages

Parallel Processing Previous Year Question

dser

Uploaded by

naj29966
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
75 views11 pages

Parallel Processing Previous Year Question

dser

Uploaded by

naj29966
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
You are on page 1/ 11
b) °° b) °°) ») 9 University of Rajshahi Department of Computer Science & Engineering B.Sc. Engineering Part IV Odd Semester Examination 2020 Course Code: CSE 4111 Course Title: Parallel Processing and Distributed System Full Marks: 52.5, Time: 3 Hours 4 [Answer any SIX (06) questions taking THREE (03) from each section] ‘ ‘ Section A Why are we developing parallel systems instead of sequential systems? Bae Why do we need to write parallel programs? 275 How can we exploit instruction level parallelism and loop level parallelism? ao : i Ifwe fetch a cache line from main memory, where in the cache should it be placed?, 3.25 What is the bottleneck of Von Neumann architecture? oof Explain which function is thread safe and which is not. cae What are the limitations of MPI_Scatter and MPI_Gather functions? **,2.50. What are the key characteristics of a vector processor? 25 Does false sharing generate incorrect result? Explain the effects of false sharing. ° ~ 3. What is cache coherence problem? Write a program to illustrate the cache 3, . coherence problem. i "0 b) 9 b) 9 How can the receiver find out the following values in MPI? , {3.95 1. The amount of data in the message. a 2. The sender of the message. 3. The tag of the message. Why do we need MPI_Reduce and MPI Allreduce function? Show the communication pattern of MPI_Allreduce. : : Write short notes on GPU. 2 Seciion B ; Why do we use MPI_Ssend instead of MPI_Send? 2 A set of numbers (15, 11, 9, 16, 3, 14, 8 7, 4, 6, 12, 10, 5, 2, 13, 1) and fout 3,25 processes are given. Show how the processes can sort the numbers using parallel odd-even transposition sort. Write a program in MPI to sort a set of numbers using -parallel odd-even 3.50 transposition sort algorithm. 8 2 3.25 3.50 3.25 Define grain size and latency. Define different types of data dependence. What are the steps of grain packing? Explain with an example. Fis oy Can multiple threads simultaneously insert new nodes into a linked list without any problem? If not, explain how to insert the nodes and write a code segment to insert the nodes by multiple threads. : ee How can we synchronize threads? re ay What is mutex? Write a program using mutex. : i eS Page of2 3 a) What are the problems of busy-waiting? b) Define critical section and race condition. ©) fibo[0] = fibof1] = 1; # pragma omp parallel for num_threads(thread_count) for (i= 2; 1

You might also like