This is my performance rig after testing the program about 4-5 hours

 VC++ 2013 and CUDA 8.0 x64 382.05
CPU:1.65%          ChainWorks Industries beta - Version 0.9.8         [21:43:14]
--------------------------------------------------------------------------------
Algo:Skunk      | Port: 2000 | NetDiff: 15568.1            | Block:      30928 |
--------------------------------------------------------------------------------
Coin:signatum   | POW: 1250  | Estimated coins/day: 210.961                    |
--------------------------------------------------------------------------------
Solved:   0 | Diff:      2.5 | Shares: 499.5   | Rej: 0.249626% | Speed:  114.62
--------------------------------------------------------------------------------
0.EVGA GTX 1080 Ti           : Shares: 239     | Rej: 1    | Speed:   53.69MH/s|
1.MSI GTX 1070               : Shares: 124.25  | Rej: 0    | Speed:   30.16MH/s|
2.MSI GTX 1070               : Shares: 136.25  | Rej: 0.25 | Speed:   31.00MH/s|
--------------------------------------------------------------------------------
0.EVGA GTX 1080 Ti           : T:63C | F:76% | P:218W | I: 25.0 | Core:1879Mhz |
1.MSI GTX 1070               : T:66C | F:84% | P:152W | I: 25.0 | Core:1956Mhz |
2.MSI GTX 1070               : T:62C | F:74% | P:168W | I: 25.0 | Core:2002Mhz |

Stable setting, no error.
this is my setting for my gpu.
GTX 1070, PL 80, Cclock, +100, memClock +500
GTX 1080 Ti, PL 80, Cclock, +100, memClock +500

This program is very nice... more faster than sp mod 3, i will try different oc setting for better performance.
Thank you chrysophylax.