Skip to content
EndsemNov 20242019 patternBE Comp

High Performance Computing (HPC) SPPU Endsem Question Paper Nov 2024

All 8 questions from this SPPU endsem paper, as printed. Read them here or open the PDF.

Open the PDFMost repeated HPC questions
Subject code
410250
Pattern
2019 pattern
Exam
Endsem
Session
Nov 2024
Max. marks
70
Time
2½ Hours
Semester
VIII
Paper code
[6354]-498

Printed for: B.E. (Computer Engineering)

Instructions to the candidates

  1. Answer Q1 or Q2, Q3 or Q4, Q5 or Q6, Q7 or Q8.
  2. Neat diagrams must be drawn wherever necessary.

Questions

  1. Q1

    1. a)

      Explain it with the help of algorithm for all-to-all broadcast on hypercube.

      [6]
    2. b)

      Describe One-to-All Broadcast and All-to-One Reduction on Rings.

      Also asked:May 2025

      [6]
    3. c)

      Explain how non-blocking communication is finalised.

      Also asked:May 2024 ·May 2023

      [5]

  2. Q2

    1. a)

      Explain Prefix-Sum Operations on eight-node hypercube with help of diagram.

      Also asked:May 2025

      [6]
    2. b)

      Explain All-to-All Broadcast and Reduction on a Ring with help of algorithm.

      Also asked:May 2025

      [6]
    3. c)

      What is non-blocking communication? Explain isend, irecv methods.

      [5]

  3. Q3

    1. a)

      With help of example, explain Effect of Granularity on Performance.

      Also asked:Nov 2025 ·May 2023

      [6]
    2. b)

      Describe a situation where fine-grained parallelism might be more appropriate than coarse- grained parallelism for improving scalability.

      [6]
    3. c)

      Describe Sources of Overhead in Parallel Programs and comments on how to avoid it.

      Also asked:Nov 2025

      [5]

  4. Q4

    1. a)

      What are the key factors that can limit the scalability of a parallel application, and how can they be addressed?

      [6]
    2. b)

      Explain Isoefficiency Metric of Scalability and Describe a scenario where an isoefficiency value of 1 would indicate perfect scalability.

      [6]
    3. c)

      Enlist performance Metrics for Parallel Systems and explain any two. 2

      Also asked:May 2024 ·May 2023

      [5]

  5. Q5

    1. a)

      Explain the concept of parallel processing in CUDA architecture and how it differs from traditional CPU-based computing.

      [8]
    2. b)

      Explain the distinction between global memory and shared memory in CUDA. How are they used differently?

      [6]
    3. c)

      Describe how communication managed and synchronized in CUDA.

      [4]

  6. Q6

    1. a)

      Explain CUDA memory model with help of CU DA memory hierarchy.

      [8]
    2. b)

      Explain processing flow of CUDA along with CUDA C functions used.

      Also asked:May 2024

      [6]
    3. c)

      Enlist and explain applications of CUDA.

      [4]

  7. Q7

    1. a)

      Explain parallel Bubble sort with help of algorithm.

      [8]
    2. b)

      Explain Kuberbets framework with help of diagram.

      [6]
    3. c)

      Explain how Document classification works in Distributed Computing.

      Also asked:May 2025

      [4]

  8. Q8

    1. a)

      Explain parallel Merge sort with help of algorithm.

      [8]
    2. b)

      Explain the concept of container orchestration in Kubernetes. How does it simplify application deployment and management?

      [6]
    3. c)

      Elaborate use ofAI/ML in Parallel Computing. 

      [4]