这就是所谓的 [Cache Oblivious Algorithm](Cache-oblivious algorithm - Wikipedia)。 在这部分的实验,我们将会尝试实现 Cache Oblivious 的一维物体热传递数值模拟算法。 一维物体的热传递可以由下面的公式给出:$$\frac{\partial u}{\partial t} = \alpha \frac{\partial^2 u}{\partial x^2}$$,其中 u (...
we develop an efficient algorithm for FA cache simulation as illustrated in Fig.1. Instead of using an O(N) or a non-linear algorithm for keeping
However, current compilers are often oblivious to cache-conscious optimizations needed to fully utilize the locality in the application. As shown later in this paper, executable binary generated by a compiler does not always fit well to the underlying cache memories. Mostly this is caused by cache...
A Cache Efficient Matrix Transpose Program? 在wikipedia 的这篇Cache-oblivious algorithm也提到了 matrix transpose的例子,甚至还画了示意图。 先看问题中的代码: for(inti=0;i<n;i++)for(intj=0;j<n;j++)destination[j+i*n]=source[i+j*n]; 这里假设 n = 4, 那么我们依次: i = 0 dst[0] =...