In Example 11.27 we suggested that there will be more cache misses than necessary if rows of the matrix Z are so long that they do not fit in the cache. If that is the case, how could you rewrite the...

In Example 11.27 we suggested that there will be more cache misses than necessary if rows of the matrix Z are so long that they do not fit in the cache. If that is the case, how could you rewrite the loop nest in order to guarantee group-spatial reuse?

Example 11.27