MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2023
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910372581081088 |
|---|---|
| author | Peng, Hongwu Xie, Xi Shivdikar, Kaustubh Hasan, MD Amit Zhao, Jiahui Huang, Shaoyi Khan, Omer Kaeli, David Ding, Caiwen |
| author_facet | Peng, Hongwu Xie, Xi Shivdikar, Kaustubh Hasan, MD Amit Zhao, Jiahui Huang, Shaoyi Khan, Omer Kaeli, David Ding, Caiwen |
| contents | In the acceleration of deep neural network training, the GPU has become the mainstream platform. GPUs face substantial challenges on GNNs, such as workload imbalance and memory access irregularities, leading to underutilized hardware. Existing solutions such as PyG, DGL with cuSPARSE, and GNNAdvisor frameworks partially address these challenges but memory traffic is still significant.
We argue that drastic performance improvements can only be achieved by the vertical optimization of algorithm and system innovations, rather than treating the speedup optimization as an "after-thought" (i.e., (i) given a GNN algorithm, designing an accelerator, or (ii) given hardware, mainly optimizing the GNN algorithm). In this paper, we present MaxK-GNN, an advanced high-performance GPU training system integrating algorithm and system innovation. (i) We introduce the MaxK nonlinearity and provide a theoretical analysis of MaxK nonlinearity as a universal approximator, and present the Compressed Balanced Sparse Row (CBSR) format, designed to store the data and index of the feature matrix after nonlinearity; (ii) We design a coalescing enhanced forward computation with row-wise product-based SpGEMM Kernel using CBSR for input feature matrix fetching and strategic placement of a sparse output accumulation buffer in shared memory; (iii) We develop an optimized backward computation with outer product-based and SSpMM Kernel.
We conduct extensive evaluations of MaxK-GNN and report the end-to-end system run-time. Experiments show that MaxK-GNN system could approach the theoretical speedup limit according to Amdahl's law. We achieve comparable accuracy to SOTA GNNs, but at a significantly increased speed: 3.22/4.24 times speedup (vs. theoretical limits, 5.52/7.27 times) on Reddit compared to DGL and GNNAdvisor implementations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2312_08656 |
| institution | arXiv |
| publishDate | 2023 |
| record_format | arxiv |
| spellingShingle | MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training Peng, Hongwu Xie, Xi Shivdikar, Kaustubh Hasan, MD Amit Zhao, Jiahui Huang, Shaoyi Khan, Omer Kaeli, David Ding, Caiwen Machine Learning Artificial Intelligence Distributed, Parallel, and Cluster Computing I.2; C.5 In the acceleration of deep neural network training, the GPU has become the mainstream platform. GPUs face substantial challenges on GNNs, such as workload imbalance and memory access irregularities, leading to underutilized hardware. Existing solutions such as PyG, DGL with cuSPARSE, and GNNAdvisor frameworks partially address these challenges but memory traffic is still significant. We argue that drastic performance improvements can only be achieved by the vertical optimization of algorithm and system innovations, rather than treating the speedup optimization as an "after-thought" (i.e., (i) given a GNN algorithm, designing an accelerator, or (ii) given hardware, mainly optimizing the GNN algorithm). In this paper, we present MaxK-GNN, an advanced high-performance GPU training system integrating algorithm and system innovation. (i) We introduce the MaxK nonlinearity and provide a theoretical analysis of MaxK nonlinearity as a universal approximator, and present the Compressed Balanced Sparse Row (CBSR) format, designed to store the data and index of the feature matrix after nonlinearity; (ii) We design a coalescing enhanced forward computation with row-wise product-based SpGEMM Kernel using CBSR for input feature matrix fetching and strategic placement of a sparse output accumulation buffer in shared memory; (iii) We develop an optimized backward computation with outer product-based and SSpMM Kernel. We conduct extensive evaluations of MaxK-GNN and report the end-to-end system run-time. Experiments show that MaxK-GNN system could approach the theoretical speedup limit according to Amdahl's law. We achieve comparable accuracy to SOTA GNNs, but at a significantly increased speed: 3.22/4.24 times speedup (vs. theoretical limits, 5.52/7.27 times) on Reddit compared to DGL and GNNAdvisor implementations. |
| title | MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training |
| topic | Machine Learning Artificial Intelligence Distributed, Parallel, and Cluster Computing I.2; C.5 |
| url | https://arxiv.org/abs/2312.08656 |