Professional Documents
Culture Documents
Q3(a) Elaborate upon synchronization and transparent scalability in CUDA GPUs. (5M)
Q3(b Write a CUDA code to add two matrices of the order N; involving kernel (5M)
) definition and launch of kernel from the host code.
Q. No. 1a 1b 2a 2b 3a 3b
2. Bring out the fundamental differences in design philosophies of CPU and GPU.
3. With a neat sketch explain the architecture of CUDA capable GPU. (state the typical
values such as computation capability, Memory, Memory bandwidth etc)
Ch2
Ch3.
3. List and explain any three CUDA APIs for managing device global memory.
5. Write a CUDA code to add two matrices of the order N; involving kernel definition and
launch of kernel from the host code.
Ch4.
Ch5
1. Give an overview of CUDA device Memory model along with CUDA device memory
types.
2. Explain the working of tile matrix multiplication kernel using shared memory along with
code.
4.