Answers Database
A1.5/F1.5 PAR - MPPR gives reasonable results at first cost table run, but unreasonable results on subsequent cost tables.
Record #4735
Problem Title:
A1.5/F1.5 PAR - MPPR gives reasonable results at first cost table run, but unreasonable
results on subsequent cost tables.
Problem Description:
When running MPPR (with 4085xl targeted) starting at cost table 2, reasonable timing scores are give
n. When MPPR moves to cost table 3 and so forth very unreasonable timing scores are given. Also ta
kes forever to run through PAR sequence. This happens in both Unix and NT.
This is a problem with the High Density Placer (XC4085XL and above).
Solution 1:
Work around.
Run PAR using only one cost table at a time. A shell script can be
created to run multiple cost tables using the "-t" switch to specify
the cost table for each pass.
This issue is still under investigation so a release date for a fix has
not been scheduled.
Solution 2:
A fix for this problem is included in the 1.5i Service Pack 1. For details
on this Service Pack see http://www.xilinx.com/techdocs/5514.htm
End of Record #4735 - Last Modified: 02/02/99 18:58 |
