Computer architecture

Question 5) We define the SIMD utilization of a program run on a GPU as the fraction of SIMD lanes that are kept busy with active threads during the run of a program.

The following code segment is run on a GPU. Each thread executes a single iteration of the shown loop. Assume that the data values of the arrays A, B, and C are already in vector registers so there are no loads and stores in this program. (Hint: Notice that there are 4 instructions in each thread.)

A warp in the GPU consists of 64 threads, and there are 64 SIMD lanes in the GPU:

for (i = 0; i < 1024768; i++) {

if (B[i] < 4444) {

A[i] = A[i] * C[i];

B[i] = A[i] + B[i];

C[i] = B[i] + 1;

}

}

(a) How many warps does it take to execute this program?

(b) When we measure the SIMD utilization for this program with one input set, we find that it is 67 / 256. What can you say about arrays A, B, and C? Be precise (Hint: Look at the if branch, what can you say about A, B and C?)

(c) Is it possible for this program to yield a SIMD utilization of 100% (circle one)? If YES, what should be true about arrays A, B, C for the SIMD utilization to be 100%? Be precise.If NO, explain why not.

(d) Is it possible for this program to yield a SIMD utilization of 25% (circle one)? If YES, what should be true about arrays A, B, and C for the SIMD utilization to be 25%? Be precise.If NO, explain why not