Parallel and Distributed Computing MCQ Practice

Question 1 of 10

A drone fleet must process 6 independent sensor files. Calibration takes 3 seconds once. Each file takes 8 seconds on one computer.

With 2 computers processing files after calibration, total time is:

  1. A. 8 seconds total, skipping calibration
  2. B. 0 seconds because files are independent
  3. C. 3 + 3*8 = 27 seconds
  4. D. 3 + 6*8 = 51 seconds only, with no benefit from the second computer

Question 2 of 10

A sequential weather model takes 90 minutes. The parallel version takes 15 minutes.

Speedup equals:

  1. A. 15/90 = 1/6
  2. B. 90+15 = 105
  3. C. 90-15 = 75
  4. D. 90/15 = 6

Question 3 of 10

A program's sequential part is 12 minutes. Its parallelizable part is 24 minutes on one core.

With 8 cores on the parallelizable part and negligible overhead, approximate total time is:

  1. A. 12 + 24/8 = 15 minutes
  2. B. 24/8 = 3 minutes, dropping the sequential part
  3. C. 12*8 + 24 = 120 minutes
  4. D. 12 + 24*8 = 204 minutes

Question 4 of 10

Which statement about increasing parallel portions is correct?

  1. A. Speedup is maximized by ignoring measured times.
  2. B. Gains are still limited by the sequential portion; eventually more parallel workers help little.
  3. C. Sequential portions shrink automatically to zero whenever a core is added.
  4. D. Parallel computing forbids any sequential steps by definition.

Question 5 of 10

Comparing efficiency of two solutions is often done by:

  1. A. Counting the number of vowels in variable names
  2. B. Ignoring runtimes entirely
  3. C. Comparing the time each takes to perform the same task
  4. D. Comparing only the font size of the source code comments

Question 6 of 10

A hospital sends a large imaging workload to 20 clinic PCs. The compute portion would be 40 minutes on one PC, but coordinating the PCs adds 10 minutes of network exchange.

Wall-clock time is about 10 + 40/20 = 12 minutes versus 40 minutes alone. Which conclusion fits?

  1. A. Speedup must be exactly 20 whenever 20 PCs exist.
  2. B. Communication overhead improves speedup beyond the sequential time.
  3. C. Distributed computing cannot use more than one device.
  4. D. Distributed computing can finish faster, yet communication means speedup is 40/12, not automatically 20.

Question 7 of 10

A maker project has N=10 independent engraving jobs of 3 minutes each and a sequential tool-calibration of 6 minutes.

Which option yields the smallest time?

  1. A. Calibrate once (6), then engrave with 5 workers so two waves of 3 minutes run: 6+6=12
  2. B. Calibrate once, then one worker does all 10 jobs: 6+30=36
  3. C. Calibrate before every job on one worker: 10*6 + 30 = 90
  4. D. Skip all engraving and only calibrate repeatedly for 60 minutes

Question 8 of 10

Parallel computing is best described as:

  1. A. Running only the sequential merge and skipping independent tasks
  2. B. Breaking work into smaller sequential operations where some run at the same time
  3. C. Never splitting a program under any circumstances
  4. D. Using smoke signals instead of processors

Question 9 of 10

A parallel schedule has sequential prep of 4 minutes and parallel tasks of 9, 7, and 5 minutes starting together.

Total time is:

  1. A. 5 minutes
  2. B. 9*7*5 minutes
  3. C. 4 + 9 = 13 minutes
  4. D. 4+9+7+5 = 25 minutes

Question 10 of 10

Why might a distributed solution with many devices fail to show efficiency gains matching the device count?

  1. A. Device counts are illegal to mention in timing discussions.
  2. B. Distributed computing forbids measuring time.
  3. C. Speedup formulas reject division on principle.
  4. D. Extra communication or coordination time can consume much of the theoretical gain.