Intel® MKL has been the most widely used math library in the data center compute industry†, holding that distinction since 2015.
This year, its capabilities have expanded.
Now known as the Intel® oneAPI Math Kernel Library (oneMKL for short), it’s been optimized for cross-architecture performance, enabling complex math-processing routines to run on CPUs, GPUs, FPGAs, and other accelerators.
Join Intel software specialist Khang Nguyen to learn how oneMKL can help you develop and deploy performant math-heavy applications, with particular focus on solving heterogeneous GPU challenges.
He’ll cover:
- oneMKL new features, including a demonstration of new sparse and LAPACK functionality, and preview of random number generator (RNG) capabilities
- Demonstrate coding examples built with Data Parallel C++ (DPC++) and oneMKL
- Demonstrate CPU-to-GPU offload capabilities
- Highlight the power of oneMKL open-source domain interfaces
- Try oneMKL (and other development tools + Intel® hardware) in the Intel® DevCloud for oneAPI
Get started with oneAPI
- Download oneMKL as part of the Intel® oneAPI Base Toolkit—includes nearly 20 dev tools and libraries for creating cross-architecture applications.
- Sign up for an Intel® DevCloud for oneAPI account—a free development sandbox with access to the latest Intel® hardware and oneAPI software.
Other resources
- Get Started with oneMKL
- Explore performance benchmarks
- Learn more about the DPC++ language
- Subscribe to the POD—Code Together is an interview series that explores the challenges at the forefront of cross-architecture development. Each bi-weekly episode features industry VIPs who are blazing new trails through today’s data-centric world. Available wherever you get your podcasts.
† Data from Evans Data Software Developer survey, 2015-2020
