Modeling Traffic of Big Data Platform for Large Scale Datacenter Networks

Xie, Zhen; Cao, Zheng; Wang, Zhan; Zang, Dawei; Shao, En; Sun, Ninghui

doi:10.1109/icpads.2016.0038

Cited by 6 publications

(2 citation statements)

References 4 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GPU for DNN training. As a prevalent high-performance architecture, GPU has been widely used for accelerating computation-intensive workloads [23,[62][63][64], particularly DNN training [8,13,25,26,42,43,55,56,56,58,66]. Mirhoseini et al [41,42] [74] introduce a pipelined structure implementing convolution and pooling layers, and train a specific DNN model LeNet on two FPGAs.…”

Section: Related Workmentioning

confidence: 99%

Enabling energy-efficient DNN training on hybrid GPU-FPGA accelerators

Liu

Xie

et al. 2021

Proceedings of the ACM International Conference on Supercomputing

Self Cite

View full text Add to dashboard Cite

DNN training consumes orders of magnitude more energy than inference and requires innovative use of accelerators to improve energy-efficiency. However, despite having complementary features, GPUs and FPGAs have been mostly used independently for the entire training process, thus neglecting the opportunity in assigning individual but distinct operations to the most suitable hardware. In this paper, we take the initiative to explore new opportunities and viable solutions in enabling energy-efficient DNN training on hybrid accelerators. To overcome fundamental challenges including avoiding training throughput loss, enabling fast design space exploration, and efficient scheduling, we propose a comprehensive framework, Hype-training, that utilizes a combination of offline characterization, performance modeling, and online scheduling of individual operations. Experimental tests using NVIDIA V100 GPUs and Intel Stratix 10 FPGAs show that, Hype-training is able to exploit a mixture of GPUs and FPGAs at a fine granularity to achieve significant energy reduction, by 44.3% on average and up to 59.7%, without any loss in training throughput. Hype-training can also enforce power caps more effectively than state-of-the-art power management mechanisms on GPUs.

show abstract

Section: Related Workmentioning

confidence: 99%

Enabling energy-efficient DNN training on hybrid GPU-FPGA accelerators

Liu

Xie

et al. 2021

Proceedings of the ACM International Conference on Supercomputing

Self Cite

View full text Add to dashboard Cite

show abstract

“…Non-volatile memory for HPC. HPC applications [15,16,28,43,67,68,70,[82][83][84] generally have significant memory consumption. A lot of work have explored the use of non-volatile memory in some HPC applications [23,53,61,80,81].…”

Section: Related Workmentioning

confidence: 99%

Md-Hm

Xie

Dong

Liu

et al. 2021

Proceedings of the ACM International Conference on Supercomputing

Self Cite

View full text Add to dashboard Cite

Molecular dynamics (MD) simulation is a fundamental method for modeling ensembles of particles. In this paper, we introduce a new method to improve the performance of MD by leveraging the emerging TB-scale big memory system. In particular, we trade memory capacity for computation capability to improve MD performance by the lookup table-based memoization technique. The traditional memoization technique for the MD simulation uses relatively small DRAM, bases on a suboptimal data structure, and replaces pair-wise computation, which leads to limited performance benefit in the big memory system. We introduce MD-HM, a memoization-based MD simulation framework customized for the big memory system. MD-HM partitions the simulation field into subgrids, and replaces computation in each subgrid as a whole based on a lightweight patternmatch algorithm to recognize computation in the subgrid. MD-HM uses a new two-phase LSM-tree to optimize read/write performance. Evaluating with nine MD simulations, we show that MD-HM outperforms the state-of-the-art LAMMPS simulation framework with an average speedup of 7.6× based on the Intel Optane-based big memory system.

show abstract

How to Build Hadoop in the Field of Transportation by Cloud Computing

2020

Data Processing Techniques and Applications for Cyber-Physical Systems (DPTA 2019)

View full text Add to dashboard Cite

Modeling Traffic of Big Data Platform for Large Scale Datacenter Networks

Cited by 6 publications

References 4 publications

Enabling energy-efficient DNN training on hybrid GPU-FPGA accelerators

Enabling energy-efficient DNN training on hybrid GPU-FPGA accelerators

Md-Hm

How to Build Hadoop in the Field of Transportation by Cloud Computing

Contact Info

Product

Resources

About