CUDA Academy · Lesson

Measuring the Speedup

Comparing naive vs tiled performance.

Prove the Win

You built a tiled kernel, but how much faster is it really? Measuring turns a guess into a number you can trust. 📊

Use CUDA events to time kernels. They sit in the GPU stream and measure exactly when work starts and finishes.

cudaEvent_t start, stop;
cudaEventCreate(&start);
cudaEventCreate(&stop);