Game Engine Architecture 6

【Game Engine Architecture 6】

1、Data-Parallel Computations

　　A GPU is a specialized coprocessor designed specifically to accelerate those computations that involve a high degree of data parallelism. It does so by combining SIMD parallelism (vectorized ALUs) with MIMD parallelism (by employing a form of preemptive multithreading). NVIDIA coined the term single instruction multiple thread (SIMT) to describe this SIMD/MIMD hybrid design, all GPUs employ the general principles of SIMT parallelism in their designs.

　　In order for a computational task to be well-suited to execution on a GPU,  the computations performed on any one element of the dataset must be independent of the results of computations on other elements.

void DotArrays_ref(unsigned count,
                    float r[],
                    const float a[],
                    const float b[])
{
    for (unsigned i = 0; i < count; ++i)
    {
        // treat each block of four floats as a
        // single four-element vector
        const unsigned j = i * 4;
        r[i] = a[j+0]*b[j+0] // ax*bx
                + a[j+1]*b[j+1] // ay*by
                + a[j+2]*b[j+2] // az*bz
                + a[j+3]*b[j+3]; // aw*bw
    }
}

View Code

　　Upper code, the computation performed by each iteration of this loop is independent of the computations performed by the other iterations.

2、Compute Kernels

　　GPGPU compute kernels are typically written in a special shading language which can be compiled into machine code that’s understandable by the GPU.

3、

4、

5、

6、