Zobrazeno 1 - 4
of 4
pro vyhledávání: '"Amarin Phaosawasdi"'
Publikováno v:
Languages and Compilers for Parallel Computing ISBN: 9783030727888
LCPC
LCPC
In a convolutional neural network (CNN), the convolution layers typically dominate the execution time. Hardware accelerators have been designed to speed up convolution. One class of accelerators provide hardware support for matrix multiplication (mat
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::19c05ed8390f238b8335e5c74a6c31fa
https://doi.org/10.1007/978-3-030-72789-5_11
https://doi.org/10.1007/978-3-030-72789-5_11
Autor:
Jeremy Johnson, Jose M. F. Moura, David Padua, André Platzer, Liang-Yan Gui, Amarin Phaosawasdi, Tze Meng Low, Soummya Kar, Manuela Veloso, Franz Franchetti, Stefan Mitsch, Juan Pablo Mendoza, Michael Franusich
Publikováno v:
IEEE Control Systems. 37:82-103
Cyberphysical systems (CPSs), ranging from critical infrastructures such as power plants, to modern (semi) autonomous vehicles, are systems that use software to control physical processes. CPSs are made up of many different computational components.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::4b56ba425ed766d160f6d1fb8910bfcd
https://doi.org/10.1109/mcs.2016.2643244
https://doi.org/10.1109/mcs.2016.2643244
Publikováno v:
WPMVP@PPoPP
Developers often rely on automatic vectorization to speed up fine-grained data-parallel code. However, for loop nests where the loops are shorter than the processor's SIMD width, automatic vectorization performs poorly. Vectorizers attempt to vectori
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::29a79fbf717fffe64e46b5b70355671d
https://doi.org/10.1145/3178433.3178436
https://doi.org/10.1145/3178433.3178436
Publikováno v:
2015 IEEE/ACM 37th IEEE International Conference on Software Engineering.
Externí odkaz:
https://explore.openaire.eu/search/publication?articleId=doi_________::38704b8eace9c470903a78a3d1c8387f
https://doi.org/10.1109/icse.2015.44
https://doi.org/10.1109/icse.2015.44