New Key features in Intel® MKL 11.0
- Conditional Numerical Reproducibility (CNR): New functionality in Intel MKL now allows you to balance performance with reproducible results by allowing greater flexibility in code branch choice and by ensuring algorithms are deterministic. See the Intel® MKL User's Guide, listen to the latest webinar, or visit this CNR article for more information. All of Intel MKL supports this new capability except the following: ScaLAPACK, cluster FFTs, data fitting functions, summary statistics functions, and the vsRngBeta random number generator on 32-bit operating systems.
- Intel MKL also introduces optimizations using the new Intel® Advanced Vector Extensions 2 (AVX2)i ncluding the new FMA3 instructions. The following parts of Intel MKL have been optimized: BLAS, FFTs, Vector math functions, Data fitting functions, Random number generators, Summary statistics functions. Please refer Intel Haswell support in Intel MKL article for more details.
- Other new Features
- LAPACK: Introduced support for LAPACK version 3.4.0
- VSL: Improved performance of viRngGeometric on Intel® Advanced Vector Extensions (AVX)
- DFT: Increased performance of 1D real-to-complex DFT on Intel® Advanced Vector Extensions (AVX)
- Removed support for Intel® Pentium® III processor. The minimal supported instruction set will be SSE2(Streaming SIMD Extensions 2)
- FFT: Completed support for real-to-complex transforms with sizes given by 64-bit integers
- Random number generators: Added support for a non-deterministic random number generator (VSL_BRNG_NONDETERM) based on the RdRand instruction
- The Intel MKL Reference Manual in HTML format is no longer available with the product. Please visit MKL Documentation page to find the latest Reference Manual.
- The Intel MKL User's Guide is available in the product, or PDFs can be browsed online here (Linux*, Windows*, Mac OS* X)
- Man pages and Eclipse help integration are no longer provided
New Key features in Intel®MKL 10.3
Intel AVX is the next step in the evolution of Intel processors. Intel AVX optimization has been extended to more MKL functions to get better performance on future Intel architecture.
- Summary Statistics library: An optimized parallel library that uses recent advances of statistics by providing modern algorithms that enhance accuracy and performance of statistical computations.
Extended MKL C language support: C interface to LAPACK and C style 0-based index arrays in PARDISO
Dynamic interface libraries for Windows: New dynamic interface libraries have been added for improved linkage from C# or Java on Windows.
Routine Level mode controls in VML: Users can now control or set the accuracy for each function separately in VML with a new argument in each function
- New symmetric matrix-vector product BLAS routine in blocked storage
- Split Complex (real real) support for 2D/3D FFTs
- New fast basic random number generator SFMT19937
- New Routine for Linear Fraction Transformation of vectors