LL-GNN: Low Latency Graph Neural Networks on FPGAs for High Energy Physics

TL;DR

This paper introduces a low latency FPGA-based graph neural network architecture tailored for high-energy physics particle detectors, achieving sub-microsecond latency and significantly improving speed and power efficiency over GPUs and prior FPGA designs.

Contribution

It proposes a novel matrix multiplication method, a fusion step, and a GNN-specific co-design approach, along with an open-source hardware template for efficient FPGA deployment under strict latency constraints.

Findings

Up to 9.0x faster than GPU implementations

Achieves up to 13.1x higher power efficiency than GPU

Latency reduced by 6.51 to 16.7x compared to previous FPGA designs

Abstract

This work presents a novel reconfigurable architecture for Low Latency Graph Neural Network (LL-GNN) designs for particle detectors, delivering unprecedented low latency performance. Incorporating FPGA-based GNNs into particle detectors presents a unique challenge since it requires sub-microsecond latency to deploy the networks for online event selection with a data rate of hundreds of terabytes per second in the Level-1 triggers at the CERN Large Hadron Collider experiments. This paper proposes a novel outer-product based matrix multiplication approach, which is enhanced by exploiting the structured adjacency matrix and a column-major data layout. Moreover, a fusion step is introduced to further reduce the end-to-end design latency by eliminating unnecessary boundaries. Furthermore, a GNN-specific algorithm-hardware co-design approach is presented which not only finds a design with a much better latency but also finds a high accuracy design under given latency constraints. To facilitate this, a customizable template for this low latency GNN hardware architecture has been designed and open-sourced, which enables the generation of low-latency FPGA designs with efficient resource utilization using a high-level synthesis tool. Evaluation results show that our FPGA implementation is up to 9.0 times faster and achieves up to 13.1 times higher power efficiency than a GPU implementation. Compared to the previous FPGA implementations, this work achieves 6.51 to 16.7 times lower latency. Moreover, the latency of our FPGA design is sufficiently low to enable deployment of GNNs in a sub-microsecond, real-time collider trigger system, enabling it to benefit from improved accuracy. The proposed LL-GNN design advances the next generation of trigger systems by enabling sophisticated algorithms to process experimental data efficiently.