430+ High Performance Computing (HPC) Solved MCQs

These multiple-choice questions (MCQs) are designed to enhance your knowledge and understanding in the following areas: Computer Science Engineering (CSE) .

Take a Test

51.	What makes a CUDA code runs in parallel
A.	__global__ indicates parallel execution of code
B.	main() function indicates parallel execution of code
C.	kernel name outside triple angle bracket indicates excecution of kernel n times in parallel
D.	first parameter value inside triple angle bracket (n) indicates excecution of kernel n times in parallel
Answer» D. first parameter value inside triple angle bracket (n) indicates excecution of kernel n times in parallel

discuss

52.	In ___________, the number of elements to be sorted is small enough to fit into the process's main memory.
A.	internal sorting
B.	internal searching
C.	external sorting
D.	external searching
Answer» A. internal sorting

discuss

53.	______________ algorithms use auxiliary storage (such as tapes and hard disks) for sorting because the number of elements to be sorted is too large to fit into memory.
A.	internal sorting
B.	internal searching
C.	external sorting
D.	external searching
Answer» C. external sorting

discuss

54.	______ can be comparison-based or noncomparison-based.
A.	searching
B.	sorting
C.	both a and b
D.	none of above
Answer» B. sorting

discuss

55.	The fundamental operation of comparison-based sorting is ________.
A.	compare-exchange
B.	searching
C.	sorting
D.	swapping
Answer» A. compare-exchange

discuss

56.	The complexity of bubble sort is Θ(n2).
A.	true
B.	false
Answer» A. true

discuss

57.	Bubble sort is difficult to parallelize since the algorithm has no concurrency.
A.	true
B.	false
Answer» A. true

discuss

58.	Quicksort is one of the most common sorting algorithms for sequential computers because of its simplicity, low overhead, and optimal average complexity.
A.	true
B.	false
Answer» A. true

discuss

59.	The performance of quicksort depends critically on the quality of the ______-.
A.	non-pivote
B.	pivot
C.	center element
D.	len of array
Answer» B. pivot

discuss

60.	the complexity of quicksort is O(nlog n).
A.	true
B.	false
Answer» A. true

discuss

61.	The main advantage of ______ is that its storage requirement is linear in the depth of the state space being searched.
A.	bfs
B.	dfs
C.	a and b
D.	none of above
Answer» B. dfs

discuss

62.	_____ algorithms use a heuristic to guide search.
A.	bfs
B.	dfs
C.	a and b
D.	none of above
Answer» A. bfs

discuss

63.	If the heuristic is admissible, the BFS finds the optimal solution.
A.	true
B.	false
Answer» A. true

discuss

64.	The search overhead factor of the parallel system is defined as the ratio of the work done by the parallel formulation to that done by the sequential formulation
A.	true
B.	false
Answer» A. true

discuss

65.	The critical issue in parallel depth-first search algorithms is the distribution of the search space among the processors.
A.	true
B.	false
Answer» A. true

discuss

66.	Graph search involves a closed list, where the major operation is a _______
A.	sorting
B.	searching
C.	lookup
D.	none of above
Answer» C. lookup

discuss

67.	Breadth First Search is equivalent to which of the traversal in the Binary Trees?
A.	pre-order traversal
B.	post-order traversal
C.	level-order traversal
D.	in-order traversal
Answer» C. level-order traversal

discuss

68.	Time Complexity of Breadth First Search is? (V – number of vertices, E – number of edges)
A.	o(v + e)
B.	o(v)
C.	o(e)
D.	o(v*e)
Answer» A. o(v + e)

discuss

69.	Which of the following is not an application of Breadth First Search?
A.	when the graph is a binary tree
B.	when the graph is a linked list
C.	when the graph is a n-ary tree
D.	when the graph is a ternary tree
Answer» B. when the graph is a linked list

discuss

70.	In BFS, how many times a node is visited?
A.	once
B.	twice
C.	equivalent to number of indegree of the node
D.	thrice
Answer» C. equivalent to number of indegree of the node

discuss

71.	Is Best First Search a searching algorithm used in graphs.
A.	true
B.	false
Answer» A. true

discuss

72.	Which of the following is not a stable sorting algorithm in its typical implementation.
A.	insertion sort
B.	merge sort
C.	quick sort
D.	bubble sort
Answer» C. quick sort

discuss ⁽¹⁾

73.	Which of the following is not true about comparison based sorting algorithms?
A.	the minimum possible time complexity of a comparison based sorting algorithm is o(nlogn) for a random input array
B.	any comparison based sorting algorithm can be made stable by using position as a criteria when two elements are compared
C.	counting sort is not a comparison based sorting algortihm
D.	heap sort is not a comparison based sorting algorithm.
Answer» D. heap sort is not a comparison based sorting algorithm.

discuss

74.	mathematically efficiency is
A.	e=s/p
B.	e=p/s
C.	e*s=p/2
D.	e=p+e/e
Answer» A. e=s/p

discuss

75.	Cost of a parallel system is sometimes referred to____ of product
A.	work
B.	processor time
C.	both
D.	none
Answer» C. both

discuss

76.	Scaling Characteristics of Parallel Programs Ts is
A.	increase
B.	constant
C.	decreases
D.	none
Answer» B. constant

discuss

77.	Speedup tends to saturate and efficiency _____ as a consequence of Amdahl’s law.
A.	increase
B.	constant
C.	decreases
D.	none
Answer» C. decreases

discuss

78.	Speedup obtained when the problem size is _______ linearly with the number of processing elements.
A.	increase
B.	constant
C.	decreases
D.	depend on problem size
Answer» A. increase

discuss

79.	The n × n matrix is partitioned among n processors, with each processor storing complete ___ of the matrix.
A.	row
B.	column
C.	both
D.	depend on processor
Answer» A. row

discuss

80.	cost-optimal parallel systems have an efficiency of ___
A.	1
B.	n
C.	logn
D.	complex
Answer» A. 1

discuss

81.	The n × n matrix is partitioned among n2 processors such that each processor owns a _____ element.
A.	n
B.	2n
C.	single
D.	double
Answer» C. single

discuss

82.	how many basic communication operations are used in matrix vector multiplication
A.	1
B.	2
C.	3
D.	4
Answer» C. 3

discuss

83.	In DNS algorithm of matrix multiplication it used
A.	1d partition
B.	2d partition
C.	3d partition
D.	both a,b
Answer» C. 3d partition

discuss

84.	In the Pipelined Execution, steps contain
A.	normalization
B.	communication
C.	elimination
D.	all
Answer» D. all

discuss

85.	the cost of the parallel algorithm is higher than the sequential run time by a factor of __
A.	2020-03-02 00:00:00
B.	2020-02-03 00:00:00
C.	3*2
D.	2/3+3/2
Answer» A. 2020-03-02 00:00:00

discuss

86.	The load imbalance problem in Parallel Gaussian Elimination: can be alleviated by using a ____ mapping
A.	acyclic
B.	cyclic
C.	both
D.	none
Answer» B. cyclic

discuss

87.	A parallel algorithm is evaluated by its runtime in function of
A.	the input size,
B.	the number of processors,
C.	the communication parameters.
D.	all
Answer» D. all

discuss

88.	For a problem consisting of W units of work, p__W processors can be used optimally.
A.	<=
B.	>=
C.	<
D.	>
Answer» A. <=

discuss

89.	C(W)__Θ(W) for optimality (necessary condition).
A.	>
B.	<
C.	<=
D.	equals
Answer» D. equals

discuss

90.	many interactions in oractical parallel programs occur in _____ pattern
A.	well defined
B.	zig-zac
C.	reverse
D.	straight
Answer» A. well defined

discuss

91.	efficient implementation of basic communication operation can improve
A.	performance
B.	communication
C.	algorithm
D.	all
Answer» A. performance

discuss

92.	efficient use of basic communication operations can reduce
A.	development effort and
B.	software quality
C.	both
D.	none
Answer» A. development effort and

discuss

93.	Group communication operations are built using_____ Messenging primitives.
A.	point-to-point
B.	one-to-all
C.	all-to-one
D.	none
Answer» A. point-to-point

discuss

94.	one processor has a piece of data and it need to send to everyone is
A.	one -to-all
B.	all-to-one
C.	point -to-point
D.	all of above
Answer» A. one -to-all

discuss

95.	the dual of one -to-all is
A.	all-to-one reduction
B.	one -to-all reduction
C.	pnoint -to-point reducntion
D.	none
Answer» A. all-to-one reduction

discuss

96.	Data items must be combined piece-wise and the result made available at
A.	target processor finally
B.	target variable finatlalyrget receiver finally
Answer» A. target processor finally

discuss

97.	wimpleat way to send p-1 messages from source to the other p-1 processors
A.	algorithm
B.	communication
C.	concurrency
D.	receiver
Answer» C. concurrency

discuss

98.	In a eight node ring, node ____ is source of broadcast
A.	1
B.	2
C.	8
D.	0
Answer» D. 0

discuss

99.	The processors compute ______ product of the vector element and the loval matrix
A.	local
B.	global
C.	both
D.	none
Answer» A. local

discuss

100.	one to all broadcast use
A.	recursive doubling
B.	simple algorithm
C.	both
D.	none
Answer» A. recursive doubling

discuss

430+ High Performance Computing (HPC) Solved MCQs

What makes a CUDA code runs in parallel

In ___________, the number of elements to be sorted is small enough to fit into the process's main memory.

______________ algorithms use auxiliary storage (such as tapes and hard disks) for sorting because the number of elements to be sorted is too large to fit into memory.

______ can be comparison-based or noncomparison-based.

The fundamental operation of comparison-based sorting is ________.

The complexity of bubble sort is Θ(n2).

Bubble sort is difficult to parallelize since the algorithm has no concurrency.

Quicksort is one of the most common sorting algorithms for sequential computers because of its simplicity, low overhead, and optimal average complexity.

The performance of quicksort depends critically on the quality of the ______-.

the complexity of quicksort is O(nlog n).

The main advantage of ______ is that its storage requirement is linear in the depth of the state space being searched.

_____ algorithms use a heuristic to guide search.

If the heuristic is admissible, the BFS finds the optimal solution.

The search overhead factor of the parallel system is defined as the ratio of the work done by the parallel formulation to that done by the sequential formulation

The critical issue in parallel depth-first search algorithms is the distribution of the search space among the processors.

Graph search involves a closed list, where the major operation is a _______

Breadth First Search is equivalent to which of the traversal in the Binary Trees?

Time Complexity of Breadth First Search is? (V – number of vertices, E – number of edges)

Which of the following is not an application of Breadth First Search?

In BFS, how many times a node is visited?

Is Best First Search a searching algorithm used in graphs.

Which of the following is not a stable sorting algorithm in its typical implementation.

Which of the following is not true about comparison based sorting algorithms?

mathematically efficiency is

Cost of a parallel system is sometimes referred to____ of product

Scaling Characteristics of Parallel Programs Ts is

Speedup tends to saturate and efficiency _____ as a consequence of Amdahl’s law.

Speedup obtained when the problem size is _______ linearly with the number of processing elements.

The n × n matrix is partitioned among n processors, with each processor storing complete ___ of the matrix.

cost-optimal parallel systems have an efficiency of ___

The n × n matrix is partitioned among n2 processors such that each processor owns a _____ element.

how many basic communication operations are used in matrix vector multiplication

In DNS algorithm of matrix multiplication it used

In the Pipelined Execution, steps contain

the cost of the parallel algorithm is higher than the sequential run time by a factor of __

The load imbalance problem in Parallel Gaussian Elimination: can be alleviated by using a ____ mapping

A parallel algorithm is evaluated by its runtime in function of

For a problem consisting of W units of work, p__W processors can be used optimally.

C(W)__Θ(W) for optimality (necessary condition).

many interactions in oractical parallel programs occur in _____ pattern

efficient implementation of basic communication operation can improve

efficient use of basic communication operations can reduce

Group communication operations are built using_____ Messenging primitives.

one processor has a piece of data and it need to send to everyone is

the dual of one -to-all is

Data items must be combined piece-wise and the result made available at

wimpleat way to send p-1 messages from source to the other p-1 processors

In a eight node ring, node ____ is source of broadcast

The processors compute ______ product of the vector element and the loval matrix

one to all broadcast use