parallel computing,ptu mtech exam,april 2012

2
M.Tech(Sem.-3rd) PARALLEL COMPUTING SUBJECT CODE : CS-517(Elective I,II,III) PAPER ID : [E0697], April 2012 Time : 03 Hours Maximum Marks:100 Instruction to Candidates: 1) Attempt any Five questions. 2) All questions carry equel marks. Q1) (a) Describe Handler’s classification for parallel computers by giving examples of each category .Also state the merits and demirts of each category (10) (b) What do you understand by systolic arrays? Give the detailed architecture of systolic arrays. Also explain the difference between systolic arrays and MIMD. (10) Q2) (a) How does data and control parallelism achieved in parallel computing? Explain with suitable examples. (10) (b) What are the laws govering performance measurement of parallel computers? Explain various performance metrices for the measurement of parallel computers with suitable examples. (10) Q3) (a) Describe various types of interconnection networks used in multiprocessor system. Also state the limitations of each. (10) (b) Discuss the issues related to the design of multiprocessor system. (10) Q4) (a) What do you understand by functional programming? Write a version of divide-and –conquer style C-Linda program that uses a fixed number of processes. (20) Q5) (a) Describe a debugging technique for shared memory parallel program with example (10)

Upload: rbakolia

Post on 24-Aug-2014

106 views

Category:

Documents


3 download

TRANSCRIPT

Page 1: Parallel Computing,PTU Mtech Exam,April 2012

M.Tech(Sem.-3rd)PARALLEL COMPUTING

SUBJECT CODE : CS-517(Elective I,II,III)PAPER ID : [E0697], April 2012

Time : 03 Hours Maximum Marks:100Instruction to Candidates:

1) Attempt any Five questions.2) All questions carry equel marks.

Q1) (a) Describe Handler’s classification for parallel computers by giving examples of each category .Also state the merits and demirts of each category (10) (b) What do you understand by systolic arrays? Give the detailed architecture of systolic arrays. Also explain the difference between systolic arrays and MIMD. (10)

Q2) (a) How does data and control parallelism achieved in parallel computing? Explain with suitable examples. (10) (b) What are the laws govering performance measurement of parallel computers? Explain various performance metrices for the measurement of parallel computers with suitable examples. (10)

Q3) (a) Describe various types of interconnection networks used in multiprocessor system. Also state the limitations of each. (10) (b) Discuss the issues related to the design of multiprocessor system. (10)

Q4) (a) What do you understand by functional programming? Write a version of divide-and –conquer style C-Linda program that uses a fixed number of processes. (20)

Q5) (a) Describe a debugging technique for shared memory parallel program with example

(10) (b) Differentiate between the following: (10) i) SIMD and MIMD ii) Shared memory and object-oriented programming.

Q6) (a) Discuss the parallelization of sequential program. (10)

(b) Which PRAM variant can best model SIMD machines and how?Explain (10)

Q7) (a) Prove that the best parallel algorithm written for an n-processor EFREWPRAM model can be no more that 0(log n) times slower that any algoritham for a CRCW model of PRAM having the same number of processors. (20)

Q8) Write short note on the followings: (20)(a) SPMD(b) Distributed memory networks(c) Loop Scheduling(d) Vector processor

Page 2: Parallel Computing,PTU Mtech Exam,April 2012