With heterogeneous computing, an essential element of optimized application performance is the ability to offload parts of the code from one hardware unit to another.
Like from a CPU to a GPU.
In this short video, technical consulting engineer Kevin O'Leary demonstrates how two Intel software tools—Intel® Advisor and Intel® VTune™ Profiler—identify the parts of CPU code that are best offloaded to a GPU and predict the latter's performance accelerating a workload by 4.4x.
Get the Software
- Download Intel® Advisor and Intel® VTune™ Profiler as part of the Intel® oneAPI Base Toolkit, a core set of tools and libraries for creating data-centric, cross-architecture applications.
- Sign up for an Intel® DevCloud account—a free development sandbox with access to the latest Intel® hardware and oneAPI software.
Technical consulting engineer, Intel Corporation