Abstract: Vectorization has been an important method of using data-level parallelism to accelerate scientific workloads on vector machines such as Cray for the past three decades. In the last decade ...
For the following IR test case, Loop vectorizer blocks vectorization of this loop due to high cost of scalar sdiv operation. Manually enabling vectorization using #pragma improves performance ...
// maximum number of while loop iterations. #define Y_STOP 10.0 void kernel_simd_class_vectorization(const Vc::double_v& x_1, const Vc::double_v& x_2, Vc::double_v& y ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results