arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2205.13675cs.ARcs.AIcs.LG
收藏

强化学习在基于数据流的粗粒度可重构阵列映射中的应用

Reinforcement Learning Approach for Mapping Applications to Dataflow-Based Coarse-Grained Reconfigurable Array

Micron Technology, Inc.(美光科技公司)

Andre Xian Ming Chang, Parth Khopkar, Bashar Romanous, Abhishek Chaurasia, Patrick Estep, Skyler Windh, Doug Vanesko, Sheik Dawood Beer Mohideen, Eugenio Culurciello

摘要:

流引擎(SE)是一个粗粒度可重配置阵列,它提供了编程灵活性和高性能与能源效率。要在SE上执行的应用程序被表示为同步数据流(SDF)图的组合,其中每个指令被表示为节点。每个节点都需要映射到SE中正确的插槽和数组,以确保程序的正确执行。这产生了具有巨大且稀疏的搜索空间的优化问题,对于该搜索空间,手动找到映射是不切实际的,因为它需要SE微架构的专业知识和知识。在这项工作中,我们提出了一个强化学习框架与全局图注意力(GGA)模块和输出屏蔽无效的位置,以发现和优化指令调度。我们使用近端策略优化来训练一个模型,该模型基于对SE设备及其约束进行建模的奖励函数将操作放置到SE瓦片中。GGA模块由图神经网络和注意力模块组成。图神经网络创建SDF的嵌入,并且注意力块用于对顺序操作布局进行建模。我们展示了某些工作负载如何映射到SE以及影响映射质量的因素的结果。我们发现,除了GGA,平均而言,发现10%更好的指令时间表的总时钟周期和掩蔽提高了20%的奖励。

英文摘要:

The Streaming Engine (SE) is a Coarse-Grained Reconfigurable Array which provides programming flexibility and high-performance with energy efficiency. An application program to be executed on the SE is represented as a combination of Synchronous Data Flow (SDF) graphs, where every instruction is represented as a node. Each node needs to be mapped to the right slot and array in the SE to ensure the correct execution of the program. This creates an optimization problem with a vast and sparse search space for which finding a mapping manually is impractical because it requires expertise and knowledge of the SE micro-architecture. In this work we propose a Reinforcement Learning framework with Global Graph Attention (GGA) module and output masking of invalid placements to find and optimize instruction schedules. We use Proximal Policy Optimization in order to train a model which places operations into the SE tiles based on a reward function that models the SE device and its constraints. The GGA module consists of a graph neural network and an attention module. The graph neural network creates embeddings of the SDFs and the attention block is used to model sequential operation placement. We show results on how certain workloads are mapped to the SE and the factors affecting mapping quality. We find that the addition of GGA, on average, finds 10% better instruction schedules in terms of total clock cycles taken and masking improves reward obtained by 20%.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑