Using CUDA\index{CUDA}, GPU kernel executions are nonblocking, and GPU/CPU data
transfers\index{CUDA!data transfer}
are blocking or nonblocking operations. All GPU kernel executions and CPU/GPU
-data transfers are associated to "streams,"\index{CUDA!stream} and all operations on a same stream
+data transfers are associated to ``streams'',\index{CUDA!stream} and all operations on a same stream
are serialized. When transferring data from the CPU to the GPU, then running GPU
computations, and finally transferring results from the GPU to the CPU, there is
a natural synchronization and serialization if these operations are achieved on
is not so generic as \Lst{algo:ch6p1overlapseqsequence}.
-\subsection{Interleaved communications-transfers-computations\\overlapping}
+\subsection{Interleaved communications-transfers-computations overlapping}
Many algorithms do not support splitting data transfers and kernel calls, and
cannot exploit CUDA streams, for example, when each GPU thread requires access to
\begin{figure}[t]
\centering
\includegraphics{Chapters/chapter6/figures/Sync-CompleteInterleaveOverlap.pdf}
- \caption{Complete overlap of internode CPU communications, CPU/GPU data transfers, and GPU
+ \caption[Complete overlap of internode CPU communications,\break\hfill CPU/GPU data transfers, and GPU
+ computations, interleaving computation-communication iterations.]{Complete overlap of internode CPU communications, CPU/GPU data transfers, and GPU
computations, interleaving computation-communication iterations.}
\label{fig:ch6p1overlapinterleaved}
\end{figure}
