As a user it is wonderful and enjoyable to live in a digitalized era, where even hand-held devices rely on heterogeneous system architecture (HSA). The computing power provided both by multiple CPUs and GPUs benifit us as users while also presenting challenges for us as developers of CFT simulation software.
Routines that free us from struggling with computing technologies using HSA and allow us to focus on innovating models and algorithms are always welcome.
This is what HILA can offer.
Developed under modern C++ standards, HILA is a lattice field theory programming framework, targeted at HPC simulations. Design features of HILA include
(1) Comprehensive class templates of commonly used mathematical and physical objects, such as matrices, which generally work as operators of a Lie algebra for non-Abelian Lie groups. A carefully implemented field class template over the lattice can accept tensors with arbitrary rank. This make writing CFT applications straightforward and intuitive through HILA's APIs.
Behind the scenes hila takes care of the MPI layout and communication. It structures the lattice fields differently for different computing platforms: 'array of structures' (standard), 'array of structures of vectors' (AVX-type), or 'structure of arrays' (GP-GPU-type).
(2) Depending on the hardware platform, optimized executables for specific hardware architectures/protocols will be produced. Details of the implementations of parallelization on different computing architecture are hidden from HILA's users, meaning that applications built with HILA's APIs can automatically run on platforms with different hardware architectures. This portability is achieved for MPI parallelization, GP-GPU parallelization as well as AVX vectorization.
The key tool for this level of potability is the HILA preprocessor (hilapp). It is a C++ source-to-source transcompiler utilizing the libtooling library of the Clang compiler. It converts HILA application C++ to platform-specific C++ code, which is passed to the appropriate platform specific compiler. This process requires Abstract Syntax Tree (AST) analysis, which is supported by libtooling. There are few simple and helpful tutorials demonstrate how this works