We thank Jonathan Owen and Joost VandeVondele for fruitful discussions. This publication is based in part on work supported by Grant No. KUS-11-009-21, made by King Abdullah University of Science and Technology (KAUST), by the Ontario Research Fund Research Excellence Program, and by the Natural Sciences and Engineering Research Council (NSERC) of Canada. Computations were performed on the BlueGene/Q supercomputer at the SciNet HPC Consortium provided through the Southern Ontario Smart Computing Innovation Platform (SOSCIP). The SOSCIP multi-university/industry consortium is funded by the Ontario Government and the Federal Economic Development Agency for Southern Ontario.