Issue No. 11 - November (1993 vol. 4)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/71.250105
<p>The GCD and Banerjee tests are the standard data dependence tests used to determinewhether a loop may be parallelized/vectorized. In an earlier work, (1991) the authorspresented a new data dependence test, the I test, which extends the accuracy of theGCD and the Banerjee tests. In the original presentation, only the case of generaldependence was considered, i.e., the case of dependence with a direction vector of theform (*,*,...,*). In the present work, the authors generalize the I test to check for datadependence subject to an arbitrary direction vector.</p>
Index Termsdirection vector I test; data dependence tests; Banerjee test; arbitrary direction vector;parallelization; GCD test; parallelism detection; algorithm theory; parallel programming
K. Psarris, X. Kong and D. Klappholz, "The Direction Vector I Test," in IEEE Transactions on Parallel & Distributed Systems, vol. 4, no. , pp. 1280-1290, 1993.