MadSpace -- Event Generation for the Era of GPUs and ML

TL;DR

MadSpace is a versatile, GPU-accelerated C++ library for phase-space and event generation, integrating neural importance sampling and machine learning workflows for high-energy physics simulations.

Contribution

It introduces a modular, unified framework with GPU support and neural importance sampling, enhancing efficiency and integration in event generation workflows.

Findings

Supports a wide range of mappings including normalizing flows

Operates on batches for end-to-end workflows

Provides seamless Python integration with ML libraries

Abstract

MadSpace is a new modular phase-space and event-generation library written in C++ with native GPU support via CUDA and HIP. It provides a unified compute-graph-based framework for phase-space construction, adaptive and neural importance sampling, and event unweighting. It includes a wide range of mappings, from the standard MadGraph multi-channel phase space to optimized normalizing flows with analytic inverse transformations. All components operate on batches of events and support end-to-end on-device workflows. A high-level Python interface enables seamless integration with machine-learning libraries such as PyTorch.