Pub Date : 2023-12-07DOI: 10.1007/s42514-023-00174-8
Zongjing Chen, Kangjin Huang, Yonggang Che, Chuanfu Xu, Jian Zhang, Z. Dai, Ming Li
{"title":"Extending OP2 framework to support portable parallel programming of complex applications","authors":"Zongjing Chen, Kangjin Huang, Yonggang Che, Chuanfu Xu, Jian Zhang, Z. Dai, Ming Li","doi":"10.1007/s42514-023-00174-8","DOIUrl":"https://doi.org/10.1007/s42514-023-00174-8","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"2 2","pages":""},"PeriodicalIF":0.9,"publicationDate":"2023-12-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"138591803","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-13DOI: 10.1007/s42514-023-00173-9
Fang Lin, Yi Liu, Xin Wang, Xueyan Gai
{"title":"Leveraging simulation of high performance computing systems with node simulation using architecture simulator","authors":"Fang Lin, Yi Liu, Xin Wang, Xueyan Gai","doi":"10.1007/s42514-023-00173-9","DOIUrl":"https://doi.org/10.1007/s42514-023-00173-9","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"23 11","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136281968","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-11-09DOI: 10.1007/s42514-023-00172-w
Shiyang Li, Jingyu Zhu, Jiaxun Han, Yuting Peng, Zhuoran Wang, Xiaoli Gong, Gang Wang, Jin Zhang, Xuqiang Wang
{"title":"OneGraph: a cross-architecture framework for large-scale graph computing on GPUs based on oneAPI","authors":"Shiyang Li, Jingyu Zhu, Jiaxun Han, Yuting Peng, Zhuoran Wang, Xiaoli Gong, Gang Wang, Jin Zhang, Xuqiang Wang","doi":"10.1007/s42514-023-00172-w","DOIUrl":"https://doi.org/10.1007/s42514-023-00172-w","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":" 10","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-11-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135241910","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Abstract NEC SX-Aurora TSUBASA (SX-AT) is the latest vector supercomputer, consisting of host processors called Vector Hosts (VHs) and vector processors called Vector Engines (VEs). The goal of this work is to simultaneously use both VHs and VEs to increase the resource utilization and improve the system throughput by co-executing more workloads. One difficulty is that performance interferences among VH and VE workloads could occur because they share some computing resources and potentially compete to use the same resource at the same time, so-called resource conflicts. To achieve efficient workload co-execution, first, this paper experimentally investigates the performance interference between a VH and a VE, when each of the two processors executes a different workload. It is empirically shown that the frequency of system calls from the VE workload could be a good indicator to predict if the co-execution could cause severe performance interference, even though monitoring system calls requires a huge runtime overhead and it is impractical to simply use it for decision making of co-execution. Then, this paper proposes a workload co-execution strategy based on a practical approach to identifying a pair of VE and VH workloads that could cause severe performance interferences. Our evaluation results clearly demonstrate that the system call frequency can be used to predict if the workload can affect the performance of another co-executing workload, and VH’s CPU load can be a good approximation of the system call frequency. The proposed approach based on the CPU loads could accurately identify a pair of workloads causing frequent resource conflicts, and thus reduce the risk of severe performance interferences between co-executing workloads on an SX-AT system, resulting in shorter makespan without significantly increasing the turn-around time.
NEC SX-Aurora TSUBASA (SX-AT)是最新的矢量超级计算机，由称为矢量主机(VHs)的主机处理器和称为矢量引擎(VEs)的矢量处理器组成。这项工作的目标是同时使用vh和ve，通过共同执行更多的工作负载来提高资源利用率和提高系统吞吐量。一个困难是，VH和VE工作负载之间可能出现性能干扰，因为它们共享一些计算资源，并可能同时竞争使用相同的资源，即所谓的资源冲突。为了实现高效的工作负载协同执行，本文首先通过实验研究了VH和VE在执行不同工作负载时的性能干扰。经验表明，来自VE工作负载的系统调用的频率可能是预测共同执行是否会导致严重性能干扰的一个很好的指标，尽管监视系统调用需要巨大的运行时开销，并且简单地将其用于共同执行的决策是不切实际的。然后，本文提出了一种基于实际方法的工作负载协同执行策略，以识别可能导致严重性能干扰的一对VE和VH工作负载。我们的评估结果清楚地表明，可以使用系统调用频率来预测工作负载是否会影响另一个协同执行工作负载的性能，并且VH的CPU负载可以很好地近似系统调用频率。所提出的基于CPU负载的方法可以准确地识别导致频繁资源冲突的一对工作负载，从而降低SX-AT系统上共同执行的工作负载之间严重性能干扰的风险，从而在不显著增加周转时间的情况下缩短makespan。
{"title":"Conflict-aware workload co-execution on SX-aurora TSUBASA","authors":"Riku Nunokawa, Yoichi Shimomura, Mulya Agung, Ryusuke Egawa, Hiroyuki Takizawa","doi":"10.1007/s42514-023-00171-x","DOIUrl":"https://doi.org/10.1007/s42514-023-00171-x","url":null,"abstract":"Abstract NEC SX-Aurora TSUBASA (SX-AT) is the latest vector supercomputer, consisting of host processors called Vector Hosts (VHs) and vector processors called Vector Engines (VEs). The goal of this work is to simultaneously use both VHs and VEs to increase the resource utilization and improve the system throughput by co-executing more workloads. One difficulty is that performance interferences among VH and VE workloads could occur because they share some computing resources and potentially compete to use the same resource at the same time, so-called resource conflicts. To achieve efficient workload co-execution, first, this paper experimentally investigates the performance interference between a VH and a VE, when each of the two processors executes a different workload. It is empirically shown that the frequency of system calls from the VE workload could be a good indicator to predict if the co-execution could cause severe performance interference, even though monitoring system calls requires a huge runtime overhead and it is impractical to simply use it for decision making of co-execution. Then, this paper proposes a workload co-execution strategy based on a practical approach to identifying a pair of VE and VH workloads that could cause severe performance interferences. Our evaluation results clearly demonstrate that the system call frequency can be used to predict if the workload can affect the performance of another co-executing workload, and VH’s CPU load can be a good approximation of the system call frequency. The proposed approach based on the CPU loads could accurately identify a pair of workloads causing frequent resource conflicts, and thus reduce the risk of severe performance interferences between co-executing workloads on an SX-AT system, resulting in shorter makespan without significantly increasing the turn-around time.","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"440 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135480691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-23DOI: 10.1007/s42514-023-00169-5
Yueyuan Zhou, ZiYi Ren, En Shao, Lixian Ma, Qiang Hu, Leping Wang, Guangming Tan
{"title":"FILL: a heterogeneous resource scheduling system addressing the low throughput problem in GROMACS","authors":"Yueyuan Zhou, ZiYi Ren, En Shao, Lixian Ma, Qiang Hu, Leping Wang, Guangming Tan","doi":"10.1007/s42514-023-00169-5","DOIUrl":"https://doi.org/10.1007/s42514-023-00169-5","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"316 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135959455","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-20DOI: 10.1007/s42514-023-00167-7
Lu Bai, Weixing Ji, Qinyuan Li, Xilai Yao, Wei Xin, Wanyi Zhu
{"title":"ConvDarts: a fast and exact convolutional algorithm selector for deep learning frameworks","authors":"Lu Bai, Weixing Ji, Qinyuan Li, Xilai Yao, Wei Xin, Wanyi Zhu","doi":"10.1007/s42514-023-00167-7","DOIUrl":"https://doi.org/10.1007/s42514-023-00167-7","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136308147","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2023-09-15DOI: 10.1007/s42514-023-00160-0
Shaojie Tan, Qingcai Jiang, Zhenwei Cao, Xiaoyu Hao, Junshi Chen, Hong An
{"title":"Uncovering the performance bottleneck of modern HPC processor with static code analyzer: a case study on Kunpeng 920","authors":"Shaojie Tan, Qingcai Jiang, Zhenwei Cao, Xiaoyu Hao, Junshi Chen, Hong An","doi":"10.1007/s42514-023-00160-0","DOIUrl":"https://doi.org/10.1007/s42514-023-00160-0","url":null,"abstract":"","PeriodicalId":29895,"journal":{"name":"CCF Transactions on High Performance Computing","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-09-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135395212","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：