OpenCores
URL https://opencores.org/ocsvn/core1990_interlaken/core1990_interlaken/trunk

Subversion Repositories core1990_interlaken

[/] [core1990_interlaken/] [trunk/] [documentation/] [protocol_survey_report/] [Sections/] [Survey_Standard.tex] - Diff between revs 5 and 9

Show entire file | Details | Blame | View Log

Rev 5 Rev 9
Line 123... Line 123...
%SPI4.2 offers advantages in channalization, contains programmable burst sizes and per-channel backpressure. Unfortunately the high width of the interface and the source-synchronous features of the protocol reduce the effective reach.
%SPI4.2 offers advantages in channalization, contains programmable burst sizes and per-channel backpressure. Unfortunately the high width of the interface and the source-synchronous features of the protocol reduce the effective reach.
 
 
 
 
\subsection{Conclusion}
\subsection{Conclusion}
        This section will conclude which of the earlier described protocols is more suitable and whether or not it meets the requirements. Table~\ref{Tab:Survey_Stardard_Overview} provides a quick overview of their specifications. In case a yes is noted, this means support is available but the exact specifications have not been described clear enough. When there is a '-' noted, there is no support or documentation has not been clear enough to provide the required information.
        This section will conclude which of the earlier described protocols is more suitable and whether or not it meets the requirements. Table~\ref{Tab:Survey_Stardard_Overview} provides a quick overview of their specifications. In case a yes is noted, this means support is available but the exact specifications have not been described clear enough. When there is a '-' noted, there is no support or documentation has not been clear enough to provide the required information.
 
        All mentioned protocols and links to their documentation can be found in Appendix \ref{Appendix:Protocol_links}.
 
 
        \taburowcolors[2] 2{tableLineOne .. tableLineTwo}
        \taburowcolors[2] 2{tableLineOne .. tableLineTwo}
        \tabulinesep = ^2mm_1mm
        \tabulinesep = ^2mm_1mm
        \everyrow{\tabucline[.3mm  white]{}}
        \everyrow{\tabucline[.3mm  white]{}}
 
 
        \begin{table}[h]
        \begin{table}[H]
        %\begin{center}
        %\begin{center}
                \begin{tabu} to \textwidth {>{\bfseries}l l l l l}
                \begin{tabu} to \textwidth {>{\bfseries}l l l l l}
                        \tableHeaderStyle
                        \tableHeaderStyle
                                                        & Interlaken      & SATA        & CPRI                   & Fibre channel \\
                                                        & Interlaken      & SATA        & CPRI                   & Fibre channel \\
                        Lane rate               & 25,3 Gbps       & 6 Gbps      & 24,33 Gbps     & 12,8 Gbps                    \\
                        Lane rate               & 25,3 Gbps       & 6 Gbps      & 24,33 Gbps     & 12,8 Gbps                    \\
Line 146... Line 147...
                \caption{Overview of the most suited protocols.}
                \caption{Overview of the most suited protocols.}
                \label{Tab:Survey_Stardard_Overview}
                \label{Tab:Survey_Stardard_Overview}
        %\end{center}
        %\end{center}
        \end{table}
        \end{table}
 
 
        SATA is an interesting protocol but the line rate is not sufficient. HyperTransport is great for huge bandwidths but implements a parallel bus while serial transmission is required in this case. Fibre channel looks like an interesting alternative. Unfortunately the lack of documentation and not being open will bring a lot of risks with it.
        SATA is an interesting protocol but the line rate is insufficient. HyperTransport is great for huge bandwidths but implements a parallel bus while serial transmission is required in this case. Fibre channel looks like an interesting alternative. Unfortunately the lack of documentation and not being open will bring a lot of risks with it.
 
 
        CPRI is another very interesting option offering a high line rate and a good way of encoding. Unfortunately the unclear documentation on CRC and flow control plus the lack of channel bonding cause this option to be a less suited option. Nevertheless a protocol to keep in mind.
        CPRI is another very interesting option offering a high line rate and a good way of encoding. Unfortunately the unclear documentation on CRC and flow control plus the lack of channel bonding cause this option to be a less suited option. Nevertheless a protocol to keep in mind.
 
 
        The Interlaken Protocol looks like the best solution. While having excellent documentation, the protocol also meets all the requirements. This even includes the optional/ nice to have specifications.
        The Interlaken Protocol comes out best. While having excellent documentation, the protocol also meets all the requirements. Even the optional/ nice to have specifications. Interlaken is open to use and is even promoted to use by Cortina Systems and Cisco Systems.
        The protocol is open to use and is even promoted to use by Cortina Systems and Cisco Systems.
 
 
 
\newpage
\newpage
 
 
 No newline at end of file
 No newline at end of file

powered by: WebSVN 2.1.0

© copyright 1999-2024 OpenCores.org, equivalent to Oliscience, all rights reserved. OpenCores®, registered trademark.