Intel MKL 10.0 Update 3 is now available

[Todd_R_Intel]

Intel MKL 10.0 Update 3 is now available. Update 3 includes the following improvements:

• Improved IA-32 version of SGEMM by 1.4 to 1.5 times for Intel Core microarchitecture.
• ZGEMM3M was sped up by up to 10 times for Intel Itanium processors and by up to 3 times for Intel Core2 Quad processors.
• The performance of factorization routines *GETRF, *POTRF, *GEQRF, SVD (bi-diagonalization routine) and the Symmetric Banded Eigensolver (tridiagonalization routine) have been improved significantly on multi-core.
• The Mac OS* X version of Intel MKL now supports Mac OS* X 10.5 Leopard.
• The LP64 interface of DSS/PARDISO now uses 64-bit addressing for internal arrays on 64-bit operating systems. This allows the direct solver to solve larger systems.
• In cases where the number of threads is set higher than the number of physical cores available and MKL_DYNAMIC = TRUE (the default value), Intel MKL will now reduce the number of threads used to be equal to the number of physical cores. For example, on systems with Hyper-Threading Technology where the OS will report the number of logical processors which is larger than the number physical processors, Intel MKL will now only use as many threads as there are physical processors or cores.
• In the Windows* and Linux* versions: A hybrid mode has been added to the MP LINPACK benchmark which can improve performance in instances where OpenMP is used on each node.
• In the Windows* and Linux* versions: Intel MKL will now set the number of threads to 1 when the MPI environment cannot be determined to be thread-safe and MKL_DYNAMIC = TRUE (the default value). See the User Guide for further information.
• In the Linux* version: Threading runtime libraries (libguide, libiomp) are now fully built with position independent code.
• Bug fixes. See the release notes for details.

To download, please login at the Intel Registration Center. Intel MKL 10.0 Update 3 is available for Linux*, Windows*, and Mac OS* X.