Cascade Pipeline for Leading-Order Matrix Element Evaluation on AMD Versal AI Engine Arrays

TL;DR

This paper introduces a cascade pipeline architecture for efficient evaluation of leading-order matrix elements on AMD Versal AI Engine arrays, achieving significant speedup and energy efficiency improvements.

Contribution

It presents a novel cascade pipeline design tailored for AMD Versal AI Engine arrays, enabling high-throughput and energy-efficient matrix element evaluations in high energy physics.

Findings

Achieved a throughput of 1.0 million evaluations per second.

Realized a 34x speedup over a single CPU core.

Improved energy efficiency by a factor of 7.7.

Abstract

A major computational bottleneck in modern High Energy Physics event generators arises from the integration of the matrix element, which requires repeated evaluations at different phase-space points to cover all possible initial- and final-state configurations. As the Large Hadron Collider enters its High-Luminosity phase, the demand for energy-efficient acceleration is expected to exceed the limits of conventional CPU scaling, motivating the use of highly parallel computing platforms such as graphics processing units (GPUs). In this work, we present an alternative approach based on a cascade pipeline architecture for evaluating leading-order matrix elements of the \ggttg process on AMD Versal AI Engine (\aie) arrays. Due to the 16\,kB per-tile program memory constraint, the computation is decomposed into a five-stage pipeline, with stages communicating via a wavefunction-token protocol over the on-chip cascade interface. Mapping 80 independent pipelines onto the 400 \aie tiles of the VCK190 platform yields a projected throughput of $1.0\times10^6$ matrix element evaluations per second at 54.8\,W, corresponding to a $34\times$ speedup over a single CPU core and a $7.7\times$ improvement in energy efficiency. Numerical agreement with the \amcnlo double-precision reference is validated at the parts-per-million level in mean relative error.