Changeset 297 for anr/section-2.tex
- Timestamp:
- Dec 16, 2010, 2:13:23 PM (14 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
anr/section-2.tex
r237 r297 8 8 software/hardware partionning, operating system, hardware design, VHDL/Verilog modeling. 9 9 Only very few SMEs have these multiple expertises and are present on the embedded system market. 10 Furthermore, even small design services in the big companies are facing the same issue. 10 11 \begin{center}\begin{minipage}{.8\linewidth}\textit{ 11 The major objective of COACH is to provide to SMEs an open-source framework to design12 embedded systems on FPGA devices by system designers.12 The major objective of COACH is to provide to system designers, an affordable 13 open-source framework to design embedded systems on FPGA devices. 13 14 }\end{minipage}\end{center} 14 15 %Current design methodologies provide quite low-level abstraction capabilities, and … … 19 20 %%% 20 21 \parlf 21 The COACH project will leverage on the expertise gained in the field of virtual prototyping 22 with the SoCLib platform, to propose a new design flow based on a small number of architectural templates. 22 The COACH project will propose a new design flow based on a small number of architectural templates. 23 23 An architectural template is a generic, parameterized architecture, relying on a predefined library 24 24 of IP cores. … … 42 42 When this interactive, system level, design space exploration is completed (converging to 43 43 a specific mapping on a specific version of the selected architectural template), the rest of the flow 44 is fully automated: The synthesisable VHDL models for the various hardware components, as well as the binary45 code for the software running on the embedded processors, and the bit-stream to program the t he target FPGA44 is fully automated: the synthesizable VHDL models for the various hardware components, as well as the binary 45 code for the software running on the embedded processors, and the bit-stream to program the target FPGA 46 46 will be automatically generated by the COACH tools. 47 47 % … … 50 50 in a platform based design flow supporting virtual prototyping and design space exploration. 51 51 Most building blocks already exist (resulting from previous projects): the GAUT 52 or UGH synthesis tools, the MUTEKH or DNA embedded operating systems, the ASIP technology, 53 the DSX exploration tool, the MWMR hardware/software communication middleware, the BEE parallelisation tool, 54 as well as the SoCLib library of systemC simulation models. They must now be enhanced and integrated in 55 a consistent design flow. 52 or UGH synthesis tools, the DNA embedded operating systems, the ASIP technology, 53 the DSX exploration tool, the MWMR hardware/software communication middleware, the BEE parallelization tool, 54 as well as the SoCLib library of SystemC simulation models. 55 They must now be enhanced and integrated in a consistent design flow: this will 56 be done in Magillem framework thanks to the IP-XACT standard. 56 57 %The five academic laboratories worked very closely during more than one year (one monthly meeting 57 58 %in Paris from january 2009 to february 2010, to analyse the issues of interfacing and integrating … … 59 60 %%% 60 61 \parlf 61 In HPC (High Performance Computing), the targeted application is an existing application62 In HPC (High Performance Computing), the targeted application is an existing one 62 63 running on a PC. 63 64 The COACH framework helps designer to accelerate it by migrating critical parts into a 64 65 SoC embedded into an FPGA device plugged to the PC PCI/X bus. 65 66 \begin{center}\begin{minipage}{.8\linewidth}\textit{ 66 The second objective of COACH is to extend the framework to HPC.67 The second objective of COACH is to extend the framework for HPC applications. 67 68 }\end{minipage}\end{center} 68 69 This will allow SMEs to enter HPC market for the applications that are 69 70 unadapted to the current GPU based solutions. 70 71 %%% 71 \parlf72 In summary, the COACH project is clearly oriented toward industry, even if most technology building blocks73 have been previously developed by academic laboratories.74 75 76 %Finally, the key points of the proposed design flow are :77 %\begin{itemize}78 %\item79 %\textbf{System level exploration}: The application coarse grain parallelism80 %is explicitely described as a Tasks and Communication Graph (TCG).81 %A template architecture is selected, and the performances are evaluated82 %on various variant of this architecture using the SoCLib virtual protyping83 %environment. This result in a specific hardware/software partitioning.84 %This system level exploration is fully controlled by the system designer, and is driven85 %by cost, throughput, latency and power consumption criteria.86 %87 %\item88 %\textbf{High Level Synthesis}: When dedicated hardware coprocessors have been89 %identified as mandatory, they will be generated by the high level synthesis (HLS) tools.90 %The COACH framework will integrate various HLS tools, supporting the micro-architectural space91 %design exploration. Here again, the exploration criteria are cost, throughput, latency92 %and power consumption.93 %At this stage, preliminary source-level transformations and optimisations by front-end94 %tools will be required to improve the efficiency of the back-end HLS tools.95 %96 %\item97 %\textbf{Early performance evaluation}: For each point in the design space,98 %figures of merit must be available such as throughput, latency, power99 %consumption, area, memory allocation and data locality. They are evaluated100 %by reliable estimators obtained by running the actual multi-task software101 %application on the virtual prototype.102 %103 %\item104 %\textbf{Independance from the Target FPGA}: The COACH description of the system105 %(both hardware and software) should be independent of the FPGA family.106 %Every point of the design space can be implemented on any FPGA component,107 %as long as it contains the hardware ressources required by the selected architectural template.108 %Basically, COACH will support both \altera and \xilinx FPGA families.109 %\end{itemize}110 %111 112 113
Note: See TracChangeset
for help on using the changeset viewer.