显而易见,全局变量可以做同样的工作,但它们带来了大家熟知的全局性名称空间污染的问题,并会因非局部性而引起错误。
Obviously, global variables could do the same job, but they cause the familiar problems with pollution of the global namespace, and allow mistakes due to non-locality.
该框架主要利用投影技术来优化数据访问的空间局部性,并同时利用数据分层技术来解决因投影而带来的数据重叠问题。
Then a new projection-delamination technique for optimizing spatial locality is presented, and a data transformation framework based on this technique is brought forward.
该文抽象地讨论了阵的空间局部性和变换保真度。
This paper itself discusses the space locality and permutation fidelity of matrix.
应用推荐