(Almost) optimal parallel block access to range queries
Mikhail J. Atallah, Sunil Prabhakar · 2000
This guarantee is true for any number of dimensions. Subsequent to this work, Bhatia et al. [4] have proved that such a performance bound is essentially optimal for this kind of scheme, and have also extended our results to the case where the number of disks is a product of the form κ1 * κ2 * … * κt where the κts need not all be 2.