Table of Contents
Welcome to the Project Wiki
This wiki provides detailed instructions on how to build and run the project. The project leverages Intel oneAPI and oneMKL to enable high-performance, multi-threaded matrix operations, including cross-device compilation and automated memory management. These tools are integral to the implementation of algorithms like Block-Lanczos, ensuring both efficiency and scalability.
The project is designed with a focus on low-level optimizations, particularly for large-scale matrix computations, and is structured to support modular builds using CMake. Whether you're exploring memory optimizations, GPU acceleration, or multi-threaded performance, this wiki will guide you through the setup and execution process 😄.
Why oneAPI?
oneAPI is an open-source, cross-platform programming model that allows writing high-performance, portable code for CPUs, GPUs, and accelerators from multiple vendors using standard C++ with SYCL. oneAPI also comes with a large ecosystem of frameworks and libraries that streamline development, for example oneMath which is SYCL optimized math kernel library.
For installation instructions, see the Installing Intel oneAPI Framework page.
For how to build and run the project, see Building and Running the-project page.
Optionally, please also check out Project configuration page.