https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/seeking-highly-efficient-caches-scheme-for-demanding-engineering/m-p/915098#M12495 <BR />seeking highly efficient caches scheme for demanding engineering computing?<BR /><BR />HI all,<BR /><BR />To same the time of costly function evaluation, I want to explore the <BR />possibility of caching.<BR /><BR />Suppose in millions of calls to the function evaluation, there are some <BR />computations, in which only a portion of the parameters change, others <BR />remain the same as the parameters that are used in some other function <BR />evaluations some time before. To save time, I can group the varying <BR />parameters into sub-groups, and cache the results for later use. This needs <BR />a highly efficient cache and a efficient organzation and look-up. Also, the <BR />decision of whether a group of parameters has been seen before should be <BR />based on data trunk in binary form, instead of decimal formats, so I can <BR />compare memory data trunks directly using memory comparison.<BR /><BR />Does anybody have good suggestions and pointers on this approach? What can Intel's libraries help me do?<BR /><BR />Thanks!<BR /><BR /><BR /><BR /><BR /><BR /><BR /><BR /> Fri, 27 Jul 2007 05:07:55 GMT losemind 2007-07-27T05:07:55Z https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/seeking-highly-efficient-caches-scheme-for-demanding-engineering/m-p/915098#M12495 <BR />seeking highly efficient caches scheme for demanding engineering computing?<BR /><BR />HI all,<BR /><BR />To same the time of costly function evaluation, I want to explore the <BR />possibility of caching.<BR /><BR />Suppose in millions of calls to the function evaluation, there are some <BR />computations, in which only a portion of the parameters change, others <BR />remain the same as the parameters that are used in some other function <BR />evaluations some time before. To save time, I can group the varying <BR />parameters into sub-groups, and cache the results for later use. This needs <BR />a highly efficient cache and a efficient organzation and look-up. Also, the <BR />decision of whether a group of parameters has been seen before should be <BR />based on data trunk in binary form, instead of decimal formats, so I can <BR />compare memory data trunks directly using memory comparison.<BR /><BR />Does anybody have good suggestions and pointers on this approach? What can Intel's libraries help me do?<BR /><BR />Thanks!<BR /><BR /><BR /><BR /><BR /><BR /><BR /><BR /> Fri, 27 Jul 2007 05:07:55 GMT https://community.intel.com/t5/Intel-oneAPI-Math-Kernel-Library/seeking-highly-efficient-caches-scheme-for-demanding-engineering/m-p/915098#M12495 losemind 2007-07-27T05:07:55Z