AP CSP Fault Tolerance & Parallel Computing (4.2, 4.3)

Which of the following is a primary benefit of making a computing system fault-tolerant?

• If one component of the system fails, users of the system can often still access it.

• If one component of the system is hacked, no information will be stolen.

• If the system becomes too expensive, making it fault-tolerant will save money.

• If the system cannot operate efficiently, making it fault-tolerant will speed up its operation.

The figure below represents a network of physically linked devices, labeled A through F. A line between two devices indicates a connection. Devices can communicate only through the connections shown.

Which of the following statements are true about the ability for devices A and C to communicate?

Select two answers.

• If devices B and D were to fail, then information sent from device A could not reach device C.

• If devices B and F were to fail, then information sent from device A could not reach device C.

• If devices D and F were to fail, then information sent from device A could not reach device C.

• If devices E and F were to fail, then information sent from device A could not reach device C.

The figure below represents a network of physically linked devices, labeled A through H. A line between two devices indicates a connection. Devices can communicate only through the connections shown.

What is the minimum number of connections that would need to be removed from the network in order for device A to not be able to communicate with device F?

• 2

• 3

• 4

• 5

Which of the following best describes the ability of parallel computing solutions to improve efficiency?

• Any problem that can be solved sequentially can be solved using a parallel solution in approximately half the time.

• Any solution can be broken down into smaller and smaller parallel portions, making the improvement in efficiency theoretically limitless as long as there are enough processors available.

• The efficiency of parallel computing solutions is rarely improved over the efficiency of sequential computing solutions.

• The efficiency of a solution that can be broken down into parallel portions is still limited by a sequential portion.

A computer has two processors that are able to run in parallel. The table below indicates the amount of time it takes either processor to execute four different processes. Assume that none of the processes is dependent on any of the other processes.

A program is used to assign processes to each of the processors. Which of the following describes how the program should assign the four processes to optimize execution time?

• Processes W and X should be assigned to one processor, and processes Y and Z should be assigned to the other processor.

• Processes W and Y should be assigned to one processor, and processes X and Z should be assigned to the other processor.

• Processes W and Z should be assigned to one processor, and processes X and Y should be assigned to the other processor.

• Process Z should be assigned to one processor, and processes W, X, and Y should be assigned to the other processor.

A certain computer has two identical processors that are able to run in parallel. The table below indicates the amount of time it takes each processor to execute each of two processes. Assume that neither process is dependent on the other.

Which of the following best approximates the difference in execution time between running the two processes in parallel instead of running them one after the other on a single processor?

• 15 seconds

• 30 seconds

• 45 seconds

• 75 seconds