Has anyone successfully set up the multithreading on Xyce? Or if there is some other nice tools for spice simulations? I am trying to run the design, but since it is only running on single core, it is taking forever.

Yes I have...

Same here. Details very much depend on the system you're compiling on. I previously failed to get Xyce to compile on an older Fedora system. After upgrading, I had better luck. I took notes while attempting the compile. What OS version are you compiling on?

Have you run the regression suite? If most of them are returning fine, it's probably not an issue with the code, but an issue with your simulation settings. That said, there is a known issue with the precompiled MPI libraries made available by default for Ubuntu 18.04. Settings wise your convergence criterion might be too strict, the netlist parsing might be being doled out in an inefficient way (

.OPTIONS DIST STRATEGY=1

for large, flat [no subcircuits] designs, and

.OPTIONS DIST STRATEGY=2

for large designs with hierarchy), and if you're doing transient simulations of circuits with combinatorial logic loops you'll need to skip the DC Operating Point check that's typically included.