false sharing

(computing theory) A performance-degrading usage pattern where the size of units in a cache means that the system may reload entire units even when it is not necessary.