The Template Task Graph (TTG) - an emerging practical dataflow programming paradigm for scientific simulation at extreme scale

Matrix multiplication is an essential part of many applications, such as linear algebra, image processing and machine learning. One platform used in such applications is TensorFlow, which is a machine learning library whose structure is based on dataflow programming paradigm. In this work, a method for multiplication of medium-density matrices on multicore CPUs using TensorFlow platform is proposed. This method, called tbt_matmul, utilizes TensorFlow built-in methods tf.matmul and tf.sparse_matmul. By partitioning each input matrix into four smaller sub-matrices, called tiles, and applying an appropriate multiplication method to each pair depending on their density, the proposed method outperforms the built-in methods for matrices of medium density and matrices of significantly uneven distribution of non-zeros.

Download Full-text

Quantification of Extreme-Scale Wind Turbine Performance Parameters due to Variations in Beam Properties

AIAA Scitech 2021 Forum ◽

10.2514/6.2021-1603 ◽

2021 ◽

Author(s):

Alejandra S. Escalera Mendoza ◽

Mayank Chetan ◽

D. Todd Griffith

Keyword(s):

Wind Turbine ◽

Performance Parameters ◽

Turbine Performance ◽

Extreme Scale

Download Full-text

Cardiovascular flow simulation at extreme scale

Computational Mechanics ◽

10.1007/s00466-009-0450-z ◽

2009 ◽

Vol 46 (1) ◽

pp. 71-82 ◽

Cited By ~ 32

Author(s):

Min Zhou ◽

Onkar Sahni ◽

H. Jin Kim ◽

C. Alberto Figueroa ◽

Charles A. Taylor ◽

...

Keyword(s):

Flow Simulation ◽

Cardiovascular Flow ◽

Extreme Scale

Download Full-text

TrainBox: An Extreme-Scale Neural Network Training Server Architecture by Systematically Balancing Operations

2020 53rd Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) ◽

10.1109/micro50266.2020.00072 ◽

2020 ◽

Author(s):

Pyeongsu Park ◽

Heetaek Jeong ◽

Jangwoo Kim

Keyword(s):

Neural Network ◽

Neural Network Training ◽

Network Training ◽

Extreme Scale ◽

Server Architecture

Download Full-text

Compact Scheduling for Task Graph Oriented Mobile Crowdsourcing

IEEE Transactions on Mobile Computing ◽

10.1109/tmc.2020.3040007 ◽

2020 ◽

pp. 1-1

Author(s):

Liang Wang ◽

Zhiwen Yu ◽

Qi Han ◽

Dingqi Yang ◽

Shirui Pan ◽

...

Keyword(s):

Task Graph ◽

Mobile Crowdsourcing

Download Full-text

PageRank Implemented with the MPI Paradigm Running on a Many-Core Neuromorphic Platform

Journal of Low Power Electronics and Applications ◽

10.3390/jlpea11020025 ◽

2021 ◽

Vol 11 (2) ◽

pp. 25

Author(s):

Evelina Forno ◽

Alessandro Salvato ◽

Enrico Macii ◽

Gianvito Urgese

Keyword(s):

Parallel Execution ◽

Spiking Neural Networks ◽

Massively Parallel ◽

Current Version ◽

Hardware Platform ◽

Pagerank Algorithm ◽

Programming Paradigm ◽

Neuromorphic Hardware ◽

Efficient Communication ◽

Many Core

SpiNNaker is a neuromorphic hardware platform, especially designed for the simulation of Spiking Neural Networks (SNNs). To this end, the platform features massively parallel computation and an efficient communication infrastructure based on the transmission of small packets. The effectiveness of SpiNNaker in the parallel execution of the PageRank (PR) algorithm has been tested by the realization of a custom SNN implementation. In this work, we propose a PageRank implementation fully realized with the MPI programming paradigm ported to the SpiNNaker platform. We compare the scalability of the proposed program with the equivalent SNN implementation, and we leverage the characteristics of the PageRank algorithm to benchmark our implementation of MPI on SpiNNaker when faced with massive communication requirements. Experimental results show that the algorithm exhibits favorable scaling for a mid-sized execution context, while highlighting that the performance of MPI-PageRank on SpiNNaker is bounded by memory size and speed limitations on the current version of the hardware.

Download Full-text