wayground logo

Free Printable Worksheets

Font size

S
M
L
XL
Worksheets

CAO M3

Total questions: 20

Worksheet time: 10mins

Name
Class
Date
1.

Which technique is used to increase ILP?

a)
  • Branch prediction

b)
  • Instruction scheduling

c)
  • Loop unrolling Loop unrolling

d)
  • All of the above

2.

Data hazards that limit ILP include:

a)
  • RAW, WAR, WAW

b)
  • Only RAW

c)
  • Only WAR

d)
  • Control hazards

3.

Which of the following is NOT a type of dependency in ILP?

a)
  • Control dependencyControl dependency

b)
  • Structural dependency

c)
  • Data dependency

d)
  • Thread dependency

4.

Pipeline stalls due to branch instructions are caused by:

a)
  • Control hazards

b)
  • Data hazards

c)
  • Structural hazards

d)
  • None

5.

Loop unrolling is done to:

a)
  • Reduce number of loop iterations

b)
  • Reduce branch penalty

c)
  • Increase available parallelism

d)
  • All of the above

6.

Software pipelining is a technique where:

a)
  • Loops are executed sequentially

b)
  • Different iterations of loop are overlapped

c)
  • Hazards are ignored

d)
  • Instructions are reordered randomly

7.

Register renaming helps to avoid:

a)
  • RAW hazards

b)
  • WAW and WAR hazards

c)
  • Structural hazards

d)
  • Control hazards

8.

Instruction scheduling at compile-time cannot handle:

a)
  • Known latencies

b)
  • Variable memory latencies

c)
  • Loop unrolling

d)
  • Static reordering

9.

A limitation of compiler-based ILP techniques is:

a)
  • They cannot detect hazards

b)
  • They rely on predictable runtime behavior

c)
  • They are slower than hardware scheduling

d)

They do not use registers

10.

Dynamic scheduling is performed by:

a)
  • Compiler

b)
  • Hardware at runtime

c)
  • Operating system

d)
  • Programmer manually

11.

The main advantage of dynamic scheduling is:

a)
  • Handling unpredictable latencies

b)
  • Lower hardware cost

c)
  • Reduced ILP

d)
  • Slower execution

12.

Tomasulo’s algorithm uses:

a)
  • Static scheduling

b)
  • Scoreboarding

c)
  • Reservation stations

d)
  • Instruction replay

13.

Dynamic scheduling can introduce:

a)
  • Deadlocks

b)
  • Pipeline stalls

c)
  • Increased hardware complexity

d)
  • All of the above

14.

Which of the following limits ILP?

a)
  • Control dependencies

b)
  • Data dependencies

c)
  • Memory aliasing

d)
  • All of the above

15.

Multithreading improves performance by:

a)
  • Hiding memory latency by switching threads

b)
  • Increasing single-thread speed

c)
  • Reducing hardware

d)
  • Eliminating cache

16.

Simultaneous multithreading (SMT) differs from coarse-grained multithreading because:

a)
  • SMT runs multiple threads in same cycle

b)
  • Coarse-grained switches only on stalls

c)
  • SMT increases ILP and TLP together

d)
  • All of the above

17.

Chip Multiprocessing (CMP) is also called:

a)
  • Multiprogramming

b)
  • Multicore

c)
  • Multiprocessing with multiple chips

d)
  • None

18.

In shared memory systems, processors communicate via:

a)
  • Message passing

b)
  • Shared variables in memory

c)
  • Network packets

d)
  • None

19.

Scalability of multicore systems depends on:

a)
  • Synchronization overhead

b)
  • Communication cost

c)
  • Parallel fraction of workload

d)
  • All of the above

20.
  1. Static scheduling works well when:

a)
  • Memory access times are highly predictable

b)
  • Memory latencies are random

c)
  • Branch mispredictions are frequent

d)
  • Cache misses dominate runtime