5. Which OpenACC is approximately equivalent to the OpenMP pragma parallel for?

6. Which OpenACC pragma is used to control data movement between the processors memory and the GPUs memory?

7. Which OpenACC clause would you use if you needed access to temporary space on the GPU?

8. The XSEDE presentation materials gives an example of a program which executes faster on 28 shared-memory cores than on a P100 GPU. What was the explanation?

9. Since each iteration of the loops must be independent, which C compiler keyword would you use to indicate that the arrays are independent?