GPU-Accelerated Sparse Matrix Vector Product based on Element-by-Element Method for Unstructured FEM using OpenACC
GPU-Accelerated Sparse Matrix Vector Product based on Element-by-Element Method for Unstructured FEM using OpenACC
复制标题
使用 OpenACC 针对非结构化 FEM 基于逐元素方法的 GPU 加速稀疏矩阵向量积
DOI：
10.1109/waccpd56842.2022.00011
复制
发表时间：
2022
期刊：
影响因子：
--
通讯作者：
Lalith Maddegedara
中科院分区：
文献类型：
--
作者：
Kusakabe Ryota;Fujita Kohei;Ichimura Tsuyoshi;Hori Muneo;Lalith Maddegedara
基于指令的并行编程模型（如OpenACC）的开发大大降低了使用GPU等加速器的成本。在这项研究中，稀疏矩阵向量积（SpMV），这往往是最昂贵的计算部分，在基于物理的模拟，是加速GPU移植使用OpenACC。通过在SpMV中引入逐元素（EBE）方法，实现了进一步的速度提升，该算法适用于GPU架构，因为它需要大量的操作但少量的内存访问。在超级计算机ABCI的一个计算节点上进行的比较中，即使使用典型的SpMV算法，使用GPU也比仅使用CPU的情况加速了22倍，并且使用EBE方法时加速了3.4倍。将该分析结果应用于考虑土壤液化的地震反应分析，与仅使用CPU相比，使用GPU的计算速度提高了42倍。
The development of directive based parallel programming models such as OpenACC has significantly reduced the cost in using accelerators such as GPUs. In this study, the sparse matrix vector product (SpMV), which was often the most computationally expensive part in physics-based simulations, was accelerated by GPU porting using OpenACC. Further speed-up was achieved by introducing the element-by-element (EBE) method in SpMV, an algorithm that is suitable for GPU architecture because it requires large amount of operations but small amount of memory access. In a comparison on one compute node of the supercomputer ABCI, using GPUs resulted in a 22- fold speedup over the CPU-only case, even when using the typical SpMV algorithm, and an additional 3.4-fold speedup when using the EBE method. The results on such analysis was applied to a seismic response analysis considering soil liquefaction, and using GPUs resulted in a 42-fold speedup compared to using only CPUs.
砂土应变空间多机制模型的三维表述及客观性
影响因子：
3.7
作者：
S. Iai
通讯作者：
S. Iai
影响因子：
3.3
作者：
Fujita, Kohei;Koyama, Kentaro;Maddegedara, Lalith
通讯作者：
Maddegedara, Lalith