【正文】
CUDA basicsa. CUDA extension to C/C++b. CUDA memory basicsc. Data transferd. A matrix multiplication examplee. Compilation and debug tools4. Practice鄧仰東第二天: 上午4. CUDA memorya. Shared memory – bank conflictb. Global memory – coalescing5. Program optimization basicsa. Basic strategiesb. Case 1: Accessing float3 typesc. Case 2: matrix transposition鄧仰東下午6. Advanced Program optimization a. Optimization strategies b. Memory coalescingc. Memory bank conflictsd. Prefe