Changeset 383 for anr/section-etat-de-art.tex
- Timestamp:
- Feb 22, 2011, 11:02:52 AM (13 years ago)
- File:
-
- 1 edited
Legend:
- Unmodified
- Added
- Removed
-
anr/section-etat-de-art.tex
r382 r383 93 93 Designers can then only generate a synthesized netlist, VHDL/Verilog simulation test 94 94 bench and custom software library that reflect the hardware configuration. 95 96 Consequently, a designer developing an embedded system needs to master four different95 \\ 96 Consequently, a designer developing an embedded system needs to master four 97 97 design environments: 98 98 \begin{enumerate} … … 156 156 would be valuable in the 157 157 context of a System Level design exploration tool. 158 \ par158 \\ 159 159 In this context, ASIP design based on Instruction Set Extensions (ISEs) has 160 160 received a lot of interest~\cite{NIOS2}, as it makes micro-architecture synthesis … … 167 167 2.5x), since ISEs performance is generally limited by I/O constraints as 168 168 they generally rely on the main CPU register file to access data. 169 170 % ( 171 %automaticcaly extraction ISE candidates for application code \cite{CODES04}, 172 %performing efficient instruction selection and/or storage resource (register) 173 %allocation \cite{FPGA08}). 169 \\ 174 170 To cope with this issue, recent approaches~\cite{DAC09,CODES08,TVLSI06} advocate the use of 175 171 micro-architectural ISE models in which the coupling between the processor micro-architecture … … 178 174 point of view and do not address the problem of generating synthesizable representations for 179 175 these models. 180 176 \\ 181 177 We therefore strongly believe that there is a need for an open-framework which 182 178 would allow researchers and system designers to : … … 201 197 data structures. Dependences (exact or conservative) are checked to guarantee 202 198 the legality of the transformation. 203 199 \\ 204 200 This has lead to the invention of many loop transformations (loop fusion, 205 201 loop splitting, loop skewing, loop interchange, loop unrolling, ...) … … 209 205 \cite{FP:96,DRV:2000}, in which the combination of two transformations is 210 206 simply a matrix product. 211 207 \\ 212 208 Since hardware is inherently parallel, finding parallelism in sequential 213 209 programs in an important prerequisite for HLS. The large FPGA chips of … … 215 211 The polyhedral model is the ideal tool for finding more parallelism in 216 212 loops. 217 213 \\ 218 214 As a side effect, it has been observed that the polyhedral model is a useful 219 215 tool for many other optimization, like memory reduction and locality
Note: See TracChangeset
for help on using the changeset viewer.