viannacl.cpp: const INTEGER Arows = 1000; const INTEGER Acols = 10000; const INTEGER Bcols = 1000; Linux 2.6.18, Intel Xeon @ 3.16GHz 6144 KB cache (unified), Telsa C1060 libscl libscl_float mult time = 15.1895 viennacl A & B copy time = 0.6607 viennacl mult time = 0.001051 viennacl C copy time = 0.633157 viennacl total time = 1.29491 GPU/CPU total time = 8.5250 per cent GPU/CPU mult time = 0.0069 per cent viennacl errors = 0 Linux 2.6.18, Intel Xeon @ 3.16GHz 6144 KB cache (unified), Telsa C1060 libsclcb libscl_float mult time = 1.69302 viennacl A & B copy time = 0.661387 viennacl mult time = 0.00103 viennacl C copy time = 0.633425 viennacl total time = 1.29584 GPU/CPU total time = 76.5401 per cent GPU/CPU mult time = 0.0608 per cent viennacl errors = 0 Mac OS X 10.6.8, Intel Core i7 @ 2.66GHz 256 KB L2 cache (per core), 4MB L3 cache GeForce GT 330M libscl libscl_float mult time = 10.753 viennacl A & B copy time = 0.655742 viennacl mult time = 0.062228 viennacl C copy time = 3.6353 viennacl total time = 4.35327 GPU/CPU total time = 40.4843 per cent GPU/CPU mult time = 0.5787 per cent viennacl errors = 0 Mac OS X 10.6.8, Intel Core i7 @ 2.66GHz 256 KB L2 cache (per core), 4MB L3 cache GeForce GT 330M libsclcb libscl_float mult time = 3.85893 viennacl A & B copy time = 0.645861 viennacl mult time = 0.062224 viennacl C copy time = 3.42285 viennacl total time = 4.13093 GPU/CPU total time = 107.0488 per cent GPU/CPU mult time = 1.6125 per cent viennacl errors = 0 Mac OS X 10.7.3, Intel Core i7 @ 2.5GHz 256 KB L2 cache (per core), 8MB L3 cache ATI Radeon HD 6770M libscl libscl_float mult time = 7.63514 viennacl A & B copy time = 0.375756 viennacl mult time = 0.107117 viennacl C copy time = 2.09576 viennacl total time = 2.57864 GPU/CPU total time = 33.7733 per cent GPU/CPU mult time = 1.4029 per cent viennacl errors = 0 Mac OS X 10.7.3, Intel Core i7 @ 2.5GHz 256 KB L2 cache (per core), 8MB L3 cache ATI Radeon HD 6770M libsclcb libscl_float mult time = 0.866822 viennacl A & B copy time = 0.380621 viennacl mult time = 0.107438 viennacl C copy time = 2.10334 viennacl total time = 2.5914 GPU/CPU total time = 298.9536 per cent GPU/CPU mult time = 12.3945 per cent viennacl errors = 0