R

Simulating Map Results

In the EXACT mode the mapping in the guide file is followed exactly. Any logic in the input NGD file that matches logic mapped into the physical components of the NCD guide file is implemented exactly as in the guide file. Mapping (including signal to pin assignments), placement and routing are all identical. Logic that is not matched to any guide component is mapped by a subsequent mapping step.

If there is a match in EXACT mode, but your constraints would conflict with the mapping in the guide file component, an error is posted. If an error is posted, you can do one of the following:

Modify the constraints to eliminate conflicts

Change to the LEVERAGE guide mode (which is less restrictive)

Modify the logical design changes to avoid conflicts

Stop using guided design

In the LEVERAGE mode, the guide design is used as a starting point in order to speed up the design process. However, in cases where the guided design tools cannot find matches or your constraints rule out any matches, the logic is not guided. Whenever the guide design conflicts with the your mapping, placement or routing constraints, the guide is ignored and your constraints are followed.

Because the LEVERAGE mode only uses the guide design as a starting point for mapping, MAP may alter the mapping to improve the speed or density of the implementation (for example, MAP may collapse additional gates into a guided CLB).

Note: Support for the leverage guide flow (–gm incremental), without a timing-driven map run of your design, specified with the map –timing option, will not be supported in future releases of Xilinx software.

For Spartan and Virtex/-E/-II/-II PRO devices, MAP uses the NGM and the NCD files as guides. You do not need to specify the NGM file on the command line. MAP infers the appropriate NGM file from the specified NCD file. If MAP does not find an NGM file in the same directory as the NCD, it generates a warning. In this case, MAP uses only the NCD file as the guide file.

Note: Guided mapping is not recommended for most HDL designs. Guided mapping depends on signal and component names, and HDL designs often have a low match rate when guided. The netlist produced after re-synthesizing HDL modules usually contains signal and instance names that are significantly different from netlists created by earlier synthesis runs. This occurs even if the source level HDL code contains only a few changes.

Simulating Map Results

When simulating with NGM files, you are not simulating a mapped result, you are simulating the logical circuit description. When simulating with NCD files, you are simulating the physical circuit description.

MAP may generate an error that is not detected in the back-annotated simulation netlist. For example, after running MAP, you can run the following command to generate the back-annotated simulation netlist:

netgen mapped.ncd mapped.ngm –o mapped.nga

Development System Reference Guide

www.xilinx.com

145

Page 145
Image 145
Xilinx 8.2i manual Simulating Map Results

8.2i specifications

Xilinx 8.2i is a significant version of the Xilinx ISE (Integrated Software Environment) that emerged in the early 2000s, marking an important milestone in the world of FPGA (Field-Programmable Gate Array) development. This version introduced a slew of advanced features, technologies, and characteristics that made it an indispensable tool for engineers and developers in designing, simulating, and implementing digital circuits.

One of the standout features of Xilinx 8.2i is its enhanced design entry capabilities. This version supports multiple design entry methods, including schematic entry, VHDL, and Verilog HDL, giving engineers the flexibility to choose their preferred approach. The integrated environment provides user-friendly graphical interfaces, making it accessible for both novice and experienced users.

Xilinx 8.2i's synthesis tools have been improved to enable more efficient design compilation and optimization. The new algorithms used in this version facilitate faster synthesis times while reducing power consumption and improving performance. Furthermore, it features support for advanced FPGA architectures, which allows for the implementation of more complex designs with greater efficiency.

The implementation tools in Xilinx 8.2i include advanced place and route capabilities, utilizing state-of-the-art algorithms for optimized resource usage. These tools enable designers to make better use of FPGA resources, ensuring that designs fit within the constraints of the target device while maximizing performance.

Another key characteristic of Xilinx 8.2i is its extensive support for various Xilinx devices such as the Spartan, Virtex, and CoolRunner series. This compatibility ensures that developers can leverage the powerful features of these FPGA families, including high-speed transceivers and DSP slices.

Xilinx 8.2i also places a strong emphasis on simulation and verification. The version integrates with various simulation tools, allowing for thorough testing of the designs before implementation. This reduces the risk of errors and ensures that the final product meets specifications.

In addition, this version includes support for design constraints, enabling engineers to specify timing, area, and other critical design parameters. By accommodating constraints, Xilinx 8.2i helps in achieving reliable and efficient designs tailored to project needs.

In summary, Xilinx 8.2i is a robust software development tool that enhances the design process for FPGAs. Its comprehensive features, including multiple design entry options, advanced synthesis and implementation tools, extensive device support, and strong simulation capabilities, make it a valuable resource for engineers and developers striving for innovation in digital design.