Practice Vectorization And Compiler Optimization (10.6) - Vector, SIMD, GPUs

Practice: Vectorization and Compiler Optimization

Lesson 10.6Vector, SIMD, GPUs

Practise this section

8 questions

Flashcard drill
2 cards from this section. Quick, good the night before a test.
Exercise
6 questions · about 4 min · pass mark 70%. Hints available.
Sign up to start the exercise
Quiz
2 questions · about 1 min · pass mark 70%. Hints available.
Sign up to start the quiz

6 sample questions

Question 1.
Define vectorization in your own words.
Show hintHide hint
Think about how data processing can be accelerated.
Question 2.
What are loop dependencies?
Show hintHide hint
Consider what would happen if the output of one loop iteration was needed in another.
3Question 3.
What is vectorization primarily used for?
- Processing single data points
- Enhancing performance by processing multiple data points
- Simply speeding up graphics
Show hintHide hint
Think about how speed improves when handling data in bulk.
4Question 4.
True or False: Compiler vectorization is always performed.
- True
- False
Show hintHide hint
Consider what might prevent a compiler from optimizing automatically.
Question 5.
Create a C++ function that performs an addition on an array of integers, first written in scalar form and then optimized using manual vectorization. Discuss the performance implications.
Show hintHide hint
Look up SIMD intrinsics for specific platforms to implement the manual vectorization.
Question 6.
Analyze a provided loop’s code for loop dependencies that prevent vectorization. Suggest modifications for optimization.
Show hintHide hint
Consider loop unrolling or redesigning data flow.

4 more questions in this section. Enroll free to practise them all.