Scribd
Upload
42 views1 page

High Performance Computing System (CSE 5154) RCS

This document contains the questions for a High Performance Computing Systems examination. It includes 3 questions: 1. Question 1 has 3 parts - 1A asks to write a parallel algorithm to perform array summation across processors and analyze time complexity. 1B explains the Rotating Daisy Chain algorithm and how it handles device priority. 1C asks to identify the appropriate MPI function when processes collect data from each other. 2. Question 2 also has 3 parts - 2A defines an OpenCL kernel and asks to write kernels to find square and add array elements. 2B writes a CUDA kernel to add matrices using one block and threads, with code to read matrices and invoke the kernel. 2C uses sample 4x3 matrices to

Uploaded by

Akshata Biradar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
42 views1 page

High Performance Computing System (CSE 5154) RCS

This document contains the questions for a High Performance Computing Systems examination. It includes 3 questions: 1. Question 1 has 3 parts - 1A asks to write a parallel algorithm to perform array summation across processors and analyze time complexity. 1B explains the Rotating Daisy Chain algorithm and how it handles device priority. 1C asks to identify the appropriate MPI function when processes collect data from each other. 2. Question 2 also has 3 parts - 2A defines an OpenCL kernel and asks to write kernels to find square and add array elements. 2B writes a CUDA kernel to add matrices using one block and threads, with code to read matrices and invoke the kernel. 2C uses sample 4x3 matrices to

Uploaded by

Akshata Biradar
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 1

I SEMESTER M.TECH.

(COMPUTER SCIENCE AND ENGINEERING)


END SEMESTER EXAMINATIONS, PART- B, FEB 2022

HIGH PERFORMANCE COMPUTING SYSTEMS [CSE 5154]


REVISED CREDIT SYSTEM
Date: 14/02/2022
Time: (75 +10) minutes MAX. MARKS: 20
_____________________________________________________________
Instructions:
 Answer ALL the questions

1A. Availability of a parallel computer system is 2-D MESH SIMD. There are N elements in
an array to be partitioned among p number of processers in mesh SIMD where p <N.
After partitioning, the respective processors perform the summation of the numbers
allocated to them simultaneously. Write a parallel algorithm to perform summation by
the respective processors after partitioning the numbers and also obtain the final sum.
Give the time complexity analysis of your algorithm. 4M

1B. With appropriate diagram, explain Rotating daisy chain algorithm. How the algorithm
will have to handle the device priority issues? 4M

1C. Identify and write the appropriate MPI function to meet the following requirement with
suitable examples:
All the processes collect data from all other processes in the same communicator,
and perform an operation on the data. 2M

2A. Define kernel for a OpenCL program. Write a kernel program to find the square of each
element of an array and add the respective elements of the original array. Also write an
equivalent code for multithreaded version for the same. 4M

2B. Write a CUDA kernel to add two Matrices A and B of dimensions M X N. Kernel uses
only one block and uses M number of CUDA threads in it. Also write code snippet of the
main program to show how do you read these two matrices and invoking the kernel to
meet the above specification? 3M

2C. In Question 2B, use two matrices each of size 4 X 3 with appropriate sample elements in
it. Explain how exactly your kernel looks at these matrices and write down the iterative
steps showing how exactly it handles the matrices. 3M

CSE 5154 Page 1 of 1

You might also like