Intel® Math Kernel Library 11.3 Update 4 Developer Guide
Automatic Offload provides performance improvements with fewer changes to the code than Compiler Assisted Offload. If you are executing a function on the host CPU, Intel MKL running in the Automatic Offload mode may offload part of the computations to one or multiple Intel Xeon Phi coprocessors without you explicitly offloading computations. By default, Intel MKL determines the best division of the work between the host CPU and coprocessors. However, you can specify a custom work division.
To enable Automatic Offload and control the division of work, use environment variables or support functions. See the Intel MKL Developer Reference for detailed descriptions of the support functions.
Use of Automatic Offload does not require changes in your link line. However, be aware that Automatic Offload supports only OpenMP* threaded Intel MKL.
|
Intel's compilers may or may not optimize to the same degree for non-Intel microprocessors for optimizations that are not unique to Intel microprocessors. These optimizations include SSE2, SSE3, and SSSE3 instruction sets and other optimizations. Intel does not guarantee the availability, functionality, or effectiveness of any optimization on microprocessors not manufactured by Intel. Microprocessor-dependent optimizations in this product are intended for use with Intel microprocessors. Certain optimizations not specific to Intel microarchitecture are reserved for Intel microprocessors. Please refer to the applicable product User and Reference Guides for more information regarding the specific instruction sets covered by this notice. Notice revision #20110804 |