I simply run
ccminer-x64.exe --benchmark -a cryptonight
My output:
*** ccminer 2.2.5 for nVidia GPUs by tpruvot@github ***
Built with VC++ 2013 and nVidia CUDA SDK 9.0 64-bits
Originally based on Christian Buchner and Christian H. project
Include some kernels from alexis78, djm34, djEzo, tsiv and krnlx.
BTC donation address: 1AJdfCpLWPNoAMDfHF1wD5y8VgKSSTHxPo (tpruvot)
[2018-05-30 20:35:47] Using JSON-RPC 2.0
[2018-05-30 20:35:48] NVML GPU monitoring enabled.
[2018-05-30 20:35:48] NVAPI GPU monitoring enabled.
[2018-05-30 20:35:48] GPU #1: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:48] 6 miner threads started, using 'cryptolight' algorithm.
[2018-05-30 20:35:48] GPU #0: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:48] GPU #2: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:48] GPU #4: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:48] GPU #5: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:48] GPU #3: Intensity set to 9, 512 threads (32x16)
[2018-05-30 20:35:59] GPU #5: ASUS GTX 1070, 819.64 H/s
[2018-05-30 20:35:59] GPU #3: Zotac GTX 1070, 791.51 H/s
[2018-05-30 20:35:59] GPU #4: EVGA GTX 1070, 813.49 H/s
[2018-05-30 20:36:08] GPU #5: ASUS GTX 1070, 793.82 H/s
[2018-05-30 20:36:08] Total: 0.00 H/s
[2018-05-30 20:36:09] GPU #3: Zotac GTX 1070, 797.62 H/s
[2018-05-30 20:36:10] GPU #4: EVGA GTX 1070, 730.74 H/s
[2018-05-30 20:36:13] GPU #5: ASUS GTX 1070, 809.25 H/s
[2018-05-30 20:36:13] Total: 0.00 H/s
[2018-05-30 20:36:13] CTRL_C_EVENT received, exiting
My specs:
Celeron G39000
4gb of ram
mobo Z270XP-SLI
Some more info:
I'm able to reproduce with 2 systems with 6x 1070s each, and 1 system with 4x 1060s
I was able to reproduce on a testing system with AMD A6 9500 (GA-AX370-Gaming 5 mobo) and different amounts/combinations of 1060s and 1070s
If I run a single card (-d 0) the "total" is correct, if I run 2 cards (-d 0,1 OR -d 0,5 OR any other combination of cards) the total becomes 0 again