Design Configuration

Abstracting Design Configuration
A high level description of the device configuration will be much easier to use in the self-checking structure and won't necessitate the interpretation of low-level bit fields.

Configuring the Design

• Once the device configuration is captured in an instance of the configuration descriptor, it will be necessary to configure the design to match.
• Grow the configuration capability. Start the simulation with the simplest configuration.

Random Design Configuration

• Maybe this is not useful in the first simulation or block-level simulation. In my opnion, it is better to use it at the last stage verification, then engineers choose the configuration randomly to verify the realiablity of the system.
  
• Based on the randomly generated system configuration, the necessary instances of the design under verification are created and connected to the necessary instances of bus-fuctional models.
  

-- My Comments The configuration file is important in the verificaton work. Because under the same entity, different architercture can be chosen. If the dimension is well devide, it is very easy to use a testbench or debugging architecture to instead the block which maybe not actually used in the verifaction.  In order to make this works, a well dimensioned system is very important. The whole chip(system), the sub-block, and sub-sub-block should be difined very clearly at beginning. Then the configuration just like a lot of MUX, choose the desired architeture that you want to embed in the testcase. This configuration somehow can speed the simulation up, because less blocks are taken into use in the testcase. Of course the simuation scripte should be written to support the valid configuration. As the book said "it may be as simple as enabling some data path, or it maybe as complicated as generating then downloading, firmware code"

FPGA/CPLD器件价格查询