Hi,
I am trying to evaluate the power consumption of one of my design. But the powerplay power analyser keeps telling me that the metric confidency is low, because much of the toggle rates are vectorless estimated (not computed based on real signal activities). Though, all the toggle rates should come from my VCD file (value change dump file) that I fill with Modelsim simulation. In the powerplay power analyser's report, I can see each signal, and wether its toggle rate has been estimated, or computed using the vcd file. Here is an exemple of what I can read :
mat_correl:matcorrel|mul_mgc_mul_pipe_1_z_oreg[0] Combinational cell Simulation file1 mat_correl:matcorrel|mul_mgc_mul_pipe_1_z_oreg[0]~feeder Combinational cell Vectorless estimation
I can trace back the first signal in my design, even in modelsim with a
find signals *pipe_1_z_oreg*
But I don't find any signal with "feeder" in the name. If someone could explain me why Quartus adds "ghost" signals with ~feeder at the end, it would help. I have the same problem with RTL or gate level simulation. Quartus 6.0 or 6.1.
many thanks in advance,
Alexandre.