Simulation with VCS has slowed down upon migration to Vivado (2014.4) based IPs
In order to enhance our standalone VCS based simulation environment, we modified all the existing Coregen based IPs to Vivado based IPs. But after this modification, we observed that the simulation was getting slowed down considerably. (ie., a simulation which gets completed in 5 min using CoregenIP was taking around than 20 minutes for completion with VivadoIP)
TOOL VERSION DETAILS: Vivado: 2014.4 VCS: vcs_mx_VI-2014.3
PROCEDURAL DETAILS: 1) The simulation library compilation was done as per procedure listed in Chapter 7 of UG900 2) The path of all compiled libraries was included in the synopsys_sim.setup file using the specific library mapping syntax. 3) The following Vivado IP definition files were also compiled together with the RTL files, - blk_mem_gen_v8_2.v - dist_mem_gen_v8_0_rfs.v - fifo_generator_v12_0_rfs.v 4) The *_funcsim.v files of the individual design IPs were compiled using a separate file list.
Once the above steps were completed, the simulation was triggered and as mentioned earlier there was a considerable increase in the run-time. Furthermore, there was no change done in the environment other than the modification of the IPs
Please provide your valuable support to fix the slow down observed in the simulation.
As per AR#59599, we tried using the UNIFAST libraries. But we were getting errors stating "Unresolved modules" which were pointing to the encrypted portions in the funcsim files.