Intel® MKL 2017 Beta ( 17 Feb 2016)
DPD200577117 | SP2DP custom library sources created and published in online article |
DPD200575663 | Fixed the problem of SVD computation of very wide matrix performing significantly slower than SVD computation of its transposition |
DPD200574694 | Introduced support for major-column format in df?Interpolate1D routines |
DPD200576806 | Fixed the Pardiso hang problem when called from an OMP critical section |
DPD200580172 | Problem of Pardiso hanging up on the 11 stage for specific matrix from Altair is now fixed |
DPD200576441 | Fixed the issue - MKL Cluster FFTW and original FFTW produce the different results in the case of out-of-place computation |
DPD200574978 | Resolved the segmentation fault resulting from a call to pthread_exit(0) |
DPD200571078 | Introduced mkl_progress support for the parallel direct sparse solver for clusters |
DPD200372223 | Improved performance of spline interpolation on multiple threads |