将量化整数Transformer部署到AMD Versal AI引擎，实现低延迟喷注分类。
Reconfigurable Computing Challenge: Transformer for Jet Tagging on Versal AI Engines
- 用可组合的AIE模块化构建Transformer，支持整数运算
- 在Versal AI引擎上完成多头注意力层映射，延迟低于100微秒
- 开源框架支持快速开发，适合高能物理实时触发研究
基于Transformer的模型在欧洲核子研究中心（CERN）大型强子对撞机（LHC）的喷注分类任务中表现优异，但将其部署于低延迟、资源受限的触发系统仍具挑战。本文首次实现了在AMD Versal AI Engine（AIE）上运行的量化整数Transformer，将密集层与多头注意力（MHA）层映射至AIE计算单元。核心贡献是一个可复用的软件框架，将Transformer层抽象为可组合的AIE构建块，并从高层Python模型描述自动生成Vitis图代码。该框架为未来研究提供基础，已开源发布于https://github.com/KastnerRG/particle_transformer_aie。
原文摘要 · Abstract (English)
Transformer-based models achieve strong performance for jet tagging at the CERN LHC, but deploying them in low-latency, resource-constrained trigger systems is challenging. We present an initial implementation of a quantized, integer-only transformer for jet tagging on the AMD Versal AI Engine (AIE), mapping dense and multi-head attention (MHA) layers to AIE tiles. The main contribution is a reusable software framework that represents transformer layers as composable AIE building blocks and automatically generates the corresponding Vitis graph code from a high-level Python model description. This framework provides a foundation for future research and is released as open-source software at https://github.com/KastnerRG/particle_transformer_aie.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。