OpenCL Runtime for OS-Driven Task Pipelining on Heterogeneous Accelerators.

Atsushi Koshiba Ryuichi Sakamoto Mitaro Namiki

Published in: RTCSA (2018)

Keyphrases

graphics processing units
operating system
transactional memory
parallel processing
data driven
parallel algorithm
field programmable gate array
fine grain
computing systems
open source
neural network
single chip
massively parallel
database manager
address space
db universal database
general purpose
data structure