一句话：提出 STeP 流式抽象，使动态张量负载能在空间数据流加速器上高效执行。
张量应用（如机器学习）中数据依赖张量形状与控制流等动态行为日益普遍，但现有面向空间数据流加速器（SDA）的编程抽象表达能力有限，迫使这些行为静态或未优化实现。论文提出 Streaming Tensor Programs（STeP）流式抽象，借助灵活路由算子、显式内存层次与符号形状语义，显式暴露动态数据速率与张量维度。由此支持动态分块、动态并行化与配置时分复用等优化，在保持数据流效率的同时适配动态行为。作者用周期近似模拟器，在代表性 LLM 层和含真实 trace 的完整模型上评估：动态分块突破先前工作的 Pareto 最优前沿，动态并行化将延迟改善约 2.72 倍，配置时分复用将计算利用率提高约 2.64 倍。
Dynamic behaviors are becoming prevalent in tensor applications, like machine learning, where many widely used models contain data-dependent tensor shapes and control flow. However, the limited expressiveness of prior programming abstractions for spatial dataflow accelerators (SDAs) forces these dynamic behaviors to be implemented statically and/or unoptimized. To address these challenges, we present Streaming Tensor Programs (STeP), a streaming abstraction that enables dynamic tensor workloads to run efficiently on SDAs. STeP introduces flexible routing operators, an explicit memory hierarchy, and symbolic-shape semantics that expose dynamic data rates and tensor dimensions. These capabilities unlock new optimizations, like dynamic tiling, dynamic parallelization, and configuration time-multiplexing, that adapt SDA execution to dynamic behaviors while preserving dataflow efficiency. Using a cycle-approximate simulator on representative LLM layers and a full model with real-world traces, STeP enables: dynamic tiling that breaks the Pareto-optimal frontier from prior work, dynamic parallelization that improves latency by ~2.72x, and configuration time-multiplexing that increases compute utilization by ~2.64x over prior SDA abstractions and their implementations.
DOI 原文 · arXiv · PDF（开放获取） · DBLP
@inproceedings{SohnZHKSZHO26,
title = {Streaming Tensor Programs: A Streaming Abstraction for Dynamic Parallelism},
author = {Gina Sohn and Genghan Zhang and Konstantin Hossfeld and Jungwoo Kim and Nathan Sobotka and Nathan Zhang and Olivia Hsu and Kunle Olukotun},
booktitle = {ASPLOS},
year = {2026},
doi = {10.1145/3779212.3790229}
}