Frankfurt Inst. for Adv. Studies, Frankfurt, Germany;
computer architecture; mainframes; parallel machines; software libraries; HPL-GPU implementation; LOEWE-CSC cluster; Linpack benchmark; accelerator cards; balanced fine-tuned workload distribution; compute nodes; computer architecture; general purpose supercomputers; heterogeneous HPC systems; heterogeneous clusters; internode heterogeneity; intranode heterogeneity; load-distributed Linpack implementation; Benchmark testing; Graphics processing units; Hardware; Niobium; Standards; Supercomputers; GPU; HPC; HPL; HPL-GPU; H;
机译:使用MPI-OpenCL在多GPU节点的群集上加速LINPACK
机译:高效节能集群上的多GPU DGEMM和高性能Linpack
机译:Linpack在GPU加速的PC群集上的计算-通信重叠
机译:用于异构簇的负载分布式的LINPACK实现
机译:致力于减轻异构集群中的同时发生的峰值功率消耗和管理能源利用
机译:群体异构数据集的进化聚类算法星的性能评估结果
机译:在CpU + GpGpU异构集群上利用混合精度算法加速Linpack性能