memory latency between a CPU and a GPU. In a CPU, tasks ``ti'' are executed one
by one with a short memory latency to get the data to process. After some tasks,
there is a context switch that allows the CPU to run concurrent applications
-and/or multi-threaded applications. {\bf REPHRASE} Memory latencies are longer in a GPU, the
+and/or multi-threaded applications. Memory latencies are longer in a GPU. Thhe
principle to obtain a high throughput is to have many tasks to
compute. Later we will see that these tasks are called threads with CUDA. With
this principle, as soon as a task is finished the next one is ready to be
-executed while the wait for data for the previous task is overlapped by
-computation of other tasks. {\bf HERE}
+executed while the wait for data for the previous task is overlapped by the
+computation of other tasks.
practice, the number of thread blocks and the size of thread blocks are given as
parameters to each kernel. Figure~\ref{ch1:fig:scalability} illustrates an
example of a kernel composed of 8 thread blocks. Then this kernel is executed on
-a small device containing only 2 SMs. {\bf RELIRE} So in this case, blocks are executed 2
+a small device containing only 2 SMs. So in this case, blocks are executed 2
by 2 in any order. If the kernel is executed on a larger CUDA device containing
4 SMs, blocks are executed 4 by 4 simultaneously. The execution times should be
approximately twice faster in the latter case. Of course, that depends on other
parameters that will be described later (in this chapter and other chapters).
-{\bf RELIRE}
-Thread blocks provide a way to cooperation in the sense that threads of the same
+
+Thread blocks provide a way to cooperate in the sense that threads of the same
block cooperatively load and store blocks of memory they all
use. Synchronizations of threads in the same block are possible (but not between
threads of different blocks). Threads of the same block can also share results
(called \texttt{blockIdx} \index{CUDA keywords!blockIdx} in CUDA) and of the
thread index (called \texttt{threadIdx}\index{CUDA keywords!threadIdx} in
CUDA). Blocks of threads and thread indexes can be decomposed into 1 dimension,
-2 dimensions, or 3 dimensions. {\bf A REGARDER} According to the dimension of manipulated data,
-the appropriate dimension can be useful. In our example, only one dimension is
+2 dimensions, or 3 dimensions. According to the dimension of manipulated data,
+the dimension of blocks of threads must be chosen carefully. In our example, only one dimension is
used. Then using the notation \texttt{.x}, we can access the first dimension
(\texttt{.y} and \texttt{.z}, respectively allow access to the second and
third dimension). The variable \texttt{blockDim}\index{CUDA keywords!blockDim}
title = "A parallel algorithm for graph matching and its MasPar implementation",
journal = "IEEE Transactions on Parallel and Distributed Systems",
volume = "8",
+number = "5",
+pages="490-501",
year = "1997"
}
author = {T. Carneiro and A. E. Muritibab and M. Negreirosc and G. A. Lima de Campos},
title = {A New Parallel Schema for Branch-and-Bound Algorithms Using {GPGPU}},
booktitle = {23rd International Symposium on Computer Architecture and High Performance Computing (SBAC-PAD)},
+pages="41-47",
+ address="New York, USA",
year = {2011}
}
author = {T. Han and T. S. Abdelrahman},
title = {Reducing branch divergence in {GPU} programs},
booktitle = {{Proceedings of the Fourth Workshop on General Purpose Processing on Graphics Processing Units (GPGPU-4), ACM}},
+ pages="1-8",
year = {2011},
publisher = {New York, USA}
}
TITLE ="An Extension of {J}ohnson's results on Job-Lot Scheduling",
JOURNAL ="Naval Research Logistis Quarterly",
YEAR ="1956",
- NOTE ="3:3"
+ pages="61-68",
+ volume ="3",
+number="3",
}
@ARTICLE{ch8:LGMitten_1959,
AUTHOR ="L. G. Mitten",
TITLE ="Sequencing $n$ jobs on two machines with arbitrary time lags",
JOURNAL ="Management Science",
+ volume="5",
+number="3",
+pages="293-298",
YEAR ="1959"
}
title = {A grid-enabled branch and bound algorithm for solving challenging combinatorial optimization problems},
booktitle = {{Proceedings of 21th IEEE International Parallel and Distributed Processing Symposium (IPDPS)}},
year = {2007},
+pages = "1-9",
month = {March},
publisher = {Long Beach, California}
}
