What is sequential computing?
A model in which operations run simultaneously on many processors
A model in which operations are performed in order, one at a time
A model in which many separate devices share one task
A model in which data is stored across the Internet
Choose your answer
Was this exam question helpful?
What is parallel computing?
Storing a program on multiple websites at once
Running every operation strictly one after another on one processor
Breaking a program into smaller operations, some of which run simultaneously on multiple processors within one system
Sending a program to users over the Internet
Choose your answer
Was this exam question helpful?
A research team runs a single climate simulation across 400 computers located in several different data centres, with each computer processing the data for one region of the world. Which computing model does this best describe?
Sequential computing, because the regions are processed one at a time
Parallel computing, because the work is split across processors inside one machine
None of these models applies to a simulation
Distributed computing, because many separate networked devices share one task
Choose your answer
Was this exam question helpful?
What mainly distinguishes distributed computing from parallel computing?
Distributed computing performs every step strictly in order
Distributed computing always uses a single processor
Parallel computing can only run on devices in different countries
Distributed computing uses multiple separate devices across a network, while parallel computing uses multiple processors within one system
Choose your answer
Was this exam question helpful?
A program has four tasks: J (8 seconds), K (4 seconds), L (2 seconds), and M (6 seconds). On two processors, J and L run on processor 1, while K and M run on processor 2. How long does the parallel solution take to finish?
10 seconds
20 seconds
12 seconds
6 seconds
Choose your answer
Was this exam question helpful?
A task takes 40 seconds to run sequentially and 16 seconds to run in parallel. What is the speedup of the parallel solution?
0.4
2.5
24
2.0
Choose your answer
Was this exam question helpful?
When tasks are split across processors, how is the running time of the parallel portion determined?
It equals the average of the processor times
It equals the time taken by the shortest-running processor
It equals the sum of the times of all processors
It equals the time taken by the longest-running processor
Choose your answer
Was this exam question helpful?
A parallel solution has a speedup of exactly 1 compared with the sequential solution. What does this mean?
The parallel solution is no faster than the sequential solution
The parallel solution is twice as fast as the sequential solution
The parallel solution takes twice as long as the sequential solution
The parallel solution failed to run
Choose your answer
Was this exam question helpful?
What is a key advantage of distributed computing compared with a single computer?
It removes the need for any sequential steps in a program
It can solve problems too large for one computer's processing time or storage, and can solve large problems faster
It guarantees that no part of the program can ever fail
It makes every problem solvable in exactly one step
Choose your answer
Was this exam question helpful?
Compared with sequential solutions, parallel solutions generally:
Can never be faster than sequential solutions
Always use exactly one processor
Scale more effectively to handle larger workloads
Remove the need to combine any results
Choose your answer
Was this exam question helpful?
A parallel program has a portion that must always run sequentially. What does this sequential portion determine?
A minimum running time the solution cannot go below, no matter how many processors are added
The maximum bandwidth of the network
The number of devices allowed in a distributed system
The order in which web pages load
Choose your answer
Was this exam question helpful?
Which two problems are best suited to distributed computing? Select two answers.
A tiny calculation that finishes almost instantly on one computer
A dataset too large to fit in a single computer's storage
A huge computation that would take a single computer far too long
A task whose every step depends on the previous step and must run one at a time
Choose your answer
Was this exam question helpful?
A program has four tasks: P (7 seconds), Q (5 seconds), R (9 seconds), and S (2 seconds). On two processors, P and Q run on processor 1, while R and S run on processor 2. How long does the parallel solution take to finish?
12 seconds
11 seconds
23 seconds
9 seconds
Choose your answer
Was this exam question helpful?
A task takes 90 seconds to run sequentially and 24 seconds to run in parallel. What is the speedup of the parallel solution?
0.27
66
3.75
3.0
Choose your answer
Was this exam question helpful?
A video-encoding solution takes 300 seconds when run sequentially. A parallel version reduces this to 80 seconds, but 25 seconds of the work must always run sequentially. If the developer keeps adding more and more processors, what is the shortest time the solution could ever approach?
25 seconds
0 seconds
80 seconds
55 seconds
Choose your answer
Was this exam question helpful?
A program has three tasks: X (10 seconds), Y (3 seconds), and Z (5 seconds). Run in parallel, X runs alone on processor 1 while Y and Z run on processor 2. What is the speedup compared with the sequential solution?
2.0
2.25
1.8
1.0
Choose your answer
Was this exam question helpful?
Why does adding more and more processors eventually stop meaningfully reducing a program's running time?
Adding processors always increases the total running time
Extra processors gradually make the whole program run sequentially
Each new processor doubles the amount of work to be done
The sequential portion cannot be parallelized and sets a minimum time the solution cannot go below
Choose your answer
Was this exam question helpful?
Why does doubling the number of processors usually not halve a program's total running time?
Adding processors always corrupts the data being processed
Part of the program must run sequentially and cannot be sped up by adding processors
Processors can only ever work one at a time
Doubling processors doubles the size of the problem
Choose your answer
Was this exam question helpful?
A program has three tasks: A (14 seconds), B (3 seconds), and C (6 seconds). Run in parallel, task A runs alone on processor 1 while tasks B and C run together on processor 2. What is the speedup of the parallel solution compared with the sequential solution?
2.56
1.64
1.53
1.0
Choose your answer
Was this exam question helpful?