DPU-v2: Energy-efficient execution of irregular directed acyclic graphs
Abstract
A growing number of applications like probabilistic machine learning, sparse linear algebra, robotic navigation, etc., exhibit irregular data flow computation that can be modeled with directed acyclic graphs (DAGs). The irregularity arises from the seemingly random connections of nodes, which makes the DAG structure unsuitable for vectorization on CPU or GPU. Moreover, the nodes usually represent a small number of arithmetic operations that cannot amortize the overhead of launching tasks/kernels for each node, further posing challenges for parallel execution. To enable energy-efficient execution, this work proposes DAG processing unit (DPU) version 2, a specialized processor architecture optimized for irregular DAGs with static connectivity. It consists of a tree-structured datapath for efficient data reuse, a customized banked register file, and interconnects tuned to support irregular register accesses. DPU-v2 is utilized effectively through a targeted compiler that systematically maps operations to the datapath, minimizes register bank conflicts, and avoids pipeline hazards. Finally, a design space exploration identifies the optimal architecture configuration that minimizes the energy-delay product. This hardware-software co-optimization approach results in a speedup of , and over a state-of-the-art DAG processor ASIP, a CPU, and a GPU, respectively, while also achieving a lower energy-delay product. In this way, this work takes an important step towards enabling an embedded execution of emerging DAG workloads.
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
2 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- Fifer: Practical Acceleration of Irregular Applications on Reconfigurable Architectures cites A Hybrid Systolic-Dataflow Architecture for Inductive Matrix Algorithms
- REASON: Accelerating Probabilistic Logical Reasoning for Scalable Neuro-Symbolic Intelligence cites CogSys: Efficient and Scalable Neurosymbolic Cognition System via Algorithm-Hardware Co-Design
- Capstan: A Vector RDA for Sparsity cites SpArch: Efficient Architecture for Sparse Matrix Multiplication
- Capstan: A Vector RDA for Sparsity cites PolyGraph: Exposing the Value of Flexibility for Graph Processing Accelerators
Cited by top-tier papers2
- CogSys: Efficient and Scalable Neurosymbolic Cognition System via Algorithm-Hardware Co-DesignZishen Wan, Hanchen Yang, Ritik Raj, Che-Kai Liu et al.HPCA 2025 · 6 citations
- REASON: Accelerating Probabilistic Logical Reasoning for Scalable Neuro-Symbolic IntelligenceZishen Wan, Che-Kai Liu, Jiayi Qian, Hanchen Yang et al.HPCA 2026 · 2 citations
Builds on5
- SpArch: Efficient Architecture for Sparse Matrix MultiplicationZhekai Zhang, Hanrui Wang, Song Han, William J. DallyHPCA 2020 · 280 citations
- A Hybrid Systolic-Dataflow Architecture for Inductive Matrix AlgorithmsJian Weng, Sihao Liu, Zhengrong Wang, Vidushi Dadu et al.HPCA 2020 · 80 citations
- Fifer: Practical Acceleration of Irregular Applications on Reconfigurable ArchitecturesQuan M. Nguyen, Daniel SánchezMICRO 2021 · 60 citations
- PolyGraph: Exposing the Value of Flexibility for Graph Processing AcceleratorsVidushi Dadu, Sihao Liu, Tony NowatzkiISCA 2021 · 60 citations
- Capstan: A Vector RDA for SparsityAlexander Rucker, Matthew Vilim, Tian Zhao, Yaqi Zhang et al.MICRO 2021 · 37 citations
Related papers
- WASP: Exploiting GPU Pipeline Parallelism with Hardware-Accelerated Automatic Warp SpecializationNeal Clayton Crago, Sana Damani, Karthikeyan Sankaralingam, Stephen W. KecklerHPCA 2024 · 13 citations
- SpZip: Architectural Support for Effective Data Compression In Irregular ApplicationsYifan Yang, Joel S. Emer, Daniel SánchezISCA 2021 · 31 citations
- Differential-Matching Prefetcher for Indirect Memory AccessGelin Fu, Tian Xia, Zhongpei Luo, Ruiyang Chen et al.HPCA 2024 · 15 citations
- Pipette: Improving Core Utilization on Irregular Applications through Intra-Core Pipeline ParallelismQuan M. Nguyen, Daniel SánchezMICRO 2020 · 28 citations
- Prodigy: Improving the Memory Latency of Data-Indirect Irregular Workloads Using Hardware-Software Co-DesignNishil Talati, Kyle May, Armand Behroozi, Yichen Yang et al.HPCA 2021 · 62 citations
