unable to solve the congestion - design route failing

[UltraGreen]

Xilinx fpga ultrascale vivado 2016.2

The design is utilizing 80% of the device and earlier it used to route with =synthesis stratagy - flow alternate routability & congestoion spread logic high for implementation

now I had to create a small 4 bit counter in the top module and the design is not routing.
I tried multiple times and the result is same ( although the set up , hold and no. of overlap nodes does change with every run. )

I cannot reduce the logic anymore, please suggest some way to get a clean bitfile.

Thanks

[barry]

Have you tried playing with your constraints? Tried all the different synthesis and par strategies? What's your speed?

[UltraGreen]

Hi Barry,

The speed is 20 MHz
I have some Multicycle path constraints. Do you mean area constraints? I havn't put any area constraints.
Will try all synthesis and par strategies . Thanks.

[TrickyDicky]

20mhz? And you're struggling to route in an ultras ale? This suggests you have done really long and large combinatorial paths. It sounds like you really need to pipeline your design.

[UltraGreen]

@Tricky,
I Agree, loo much combinatorial, as the rtl was meant for ASIC . cant do much on that front. I mean cant pipeline.
Also the device utilization has gone up to 82% on this Stacked silicon device

[ftsolutions]

Have you many wide/long buses within the design? If it is only 20MHz maybe you can replace a parallel bus with a serial one? I haven't had to resort to this in my designs (yet), but this general concept may be of some use:
http://electronicdesign.com/eda/usin...e-fpga-routing