README.md

📚 funlib

funlib is a lightweight, modular C++ linear algebra library focused on efficient matrix and vector operations. It is GPU-accelerated using SYCL, making it ideal for numerical simulations and scientific computing tasks such as Finite Element Methods (FEM).

✨ Features

Matrix and Vector operations (row-major storage)

Matrix-Matrix and Matrix-Vector multiplication

SYCL-accelerated kernels

Custom Conjugate Gradient Solver for large sparse systems

Eigenvalue and Eigenvector computation (under development)

Simple CPU fallback when no GPU available

Clean and extensible design

Designed for FEM and scientific simulations

🛠 Technologies

C++17

SYCL (Intel oneAPI DPC++, or LLVM SYCL builds)

CMake (build system)

🏗️ Build Instructions

Make sure you have:

A SYCL-capable compiler (e.g., Intel oneAPI DPC++, or LLVM SYCL build)

CMake version 3.15 or higher

Clone funlib

git clone https://github.com/yourusername/funlib.git cd funlib

Create a build directory

mkdir build cd build

Configure with CMake

cmake ..

Build

make

Notes:

If you are using a manually built SYCL Clang, ensure your CMakeLists.txt points to the correct clang++ binary.

You may need to set the environment for oneAPI compilers (source /path/to/oneapi/setvars.sh).

🚀 Quick Start Example

#include "funlib.hpp"

int main()
{
    flib::Set<double> A(3, 3);
    flib::Set<double> x(3, 1);

    // Fill matrix A and vector x
    A(0,0) = 4; A(0,1) = 1; A(0,2) = 2;
    A(1,0) = 1; A(1,1) = 3; A(1,2) = 0;
    A(2,0) = 2; A(2,1) = 0; A(2,2) = 1;
    
    x(0,0) = 1;
    x(1,0) = 2;
    x(2,0) = 3;

    auto b = prod(A, x); // Matrix-vector multiplication

    b.print(); // Display result
}

📈 Roadmap

Matrix and vector classes

Matrix-matrix and matrix-vector product

SYCL GPU acceleration

Conjugate Gradient solver

Eigenvalue computation

Sparse matrix support (CSR format)

    Preconditioners for iterative solvers

📜 License

funlib is open-source under the MIT License.