FIESTA 4: optimized Feynman integral calculations with GPU support

TL;DR

FIESTA4 advances sector-decomposition-based Feynman integral evaluation by emphasizing high-scale performance, GPU acceleration, and robust parallelization. It integrates SIMD vectorization, CUDA GPU support, and MPI-based cluster execution to deliver substantial speedups over prior versions while retaining stability and a rich option set for accuracy control. The paper provides a comprehensive guide to installation, code structure, Mathematica interfaces, and pool/worker configurations, underscoring practical routes to scalable high-precision epsilon expansions. Benchmarks on four-loop diagrams illustrate significant improvements, validating the approach for large-scale, GPU-accelerated computations in high-energy physics.

Abstract

This paper presents a new major release of the program FIESTA (Feynman Integral Evaluation by a Sector decomposiTion Approach). The new release is mainly aimed at optimal performance at large scales when one is increasing the number of sampling points in order to reduce the uncertainty estimates. The release now supports graphical processor units (GPU) for the numerical integration, methods to optimize cluster-usage, as well as other speed, memory, and stability improvements.

Figures (2)

• Figure 1: Four-loop non-planar on-shell propagator diagram
• Figure 2: Massless box diagram