Quote from: crypto4pizza on February 19, 2018, 05:52:57 PM

Quote from: Shodo on February 19, 2018, 02:28:43 PM

Can someone please explain me one thing.
When I run DSTM or EWBF I have one card that starts ok but then goes down very fast to 1/2 hashrate.
NVIDIA Inspector shows very small (about 30-40%) usage of MCU on that card.
If I run ccminer (any algo) - all is fine, all cards are equal. So it seems to me that this is an equihash issue.
Please check card 1. Sometimes it shows shares like .. That card is totally equal with 7.8.9 GPUs, so it must give 750 h/s

Defective card?

2018-02-19 17:47:37| ========== Sol/s: 7020.9 Sol/W: 2.65 Avg: 7058.2 I/s: 3777.9 Sh: 3.89 1.00 184
2018-02-19 17:47:42| GPU0 53C Sol/s: 700.6 Sol/W: 3.29 Avg: 698.9 I/s: 370.6 Sh: 0.18 1.00 172
2018-02-19 17:47:43| GPU1 53C Sol/s: 371.2 Sol/W: 1.96 Avg: 370.1 I/s: 199.0 Sh: 0.00 . .
2018-02-19 17:47:45| GPU2 68C Sol/s: 756.9 Sol/W: 2.57 Avg: 755.6 I/s: 405.2 Sh: 0.54 1.00 187 +
2018-02-19 17:47:47| GPU3 64C Sol/s: 735.8 Sol/W: 2.70 Avg: 742.6 I/s: 398.2 Sh: 0.18 1.00 188
2018-02-19 17:47:48| GPU4 66C Sol/s: 761.2 Sol/W: 2.71 Avg: 759.7 I/s: 405.8 Sh: 0.36 1.00 188 +
2018-02-19 17:47:50| GPU5 61C Sol/s: 754.7 Sol/W: 2.61 Avg: 760.5 I/s: 408.4 Sh: 0.73 1.00 187
2018-02-19 17:47:53| GPU6 66C Sol/s: 690.1 Sol/W: 2.83 Avg: 694.7 I/s: 372.7 Sh: 0.36 1.00 172
2018-02-19 17:47:54| GPU7 56C Sol/s: 758.7 Sol/W: 2.50 Avg: 761.1 I/s: 407.2 Sh: 0.18 1.00 187
2018-02-19 17:47:57| GPU8 55C Sol/s: 751.9 Sol/W: 2.76 Avg: 754.9 I/s: 404.9 Sh: 0.54 1.00 187
2018-02-19 17:47:59| GPU9 62C Sol/s: 769.0 Sol/W: 2.56 Avg: 759.6 I/s: 406.4 Sh: 0.91 1.00 188
2018-02-19 17:47:59| ========== Sol/s: 7050.1 Sol/W: 2.65 Avg: 7057.6 I/s: 3778.4 Sh: 3.99 1.00 184
2018-02-19 17:48:04| GPU0 53C Sol/s: 693.4 Sol/W: 3.29 Avg: 698.6 I/s: 370.9 Sh: 0.17 1.00 172
2018-02-19 17:48:05| GPU1 54C Sol/s: 370.9 Sol/W: 1.96 Avg: 370.2 I/s: 199.0 Sh: 0.00 . .
2018-02-19 17:48:07| GPU2 69C Sol/s: 748.8 Sol/W: 2.56 Avg: 755.2 I/s: 405.2 Sh: 0.51 1.00 187
2018-02-19 17:48:09| GPU3 65C Sol/s: 742.9 Sol/W: 2.70 Avg: 742.6 I/s: 398.2 Sh: 0.17 1.00 188
2018-02-19 17:48:10| GPU4 66C Sol/s: 753.4 Sol/W: 2.71 Avg: 759.3 I/s: 405.8 Sh: 0.34 1.00 188
2018-02-19 17:48:13|> GPU5 61C Sol/s: 745.9 Sol/W: 2.60 Avg: 759.6 I/s: 408.4 Sh: 0.68 1.00 187
2018-02-19 17:48:15|> GPU6 66C Sol/s: 701.3 Sol/W: 2.83 Avg: 695.1 I/s: 373.0 Sh: 0.34 1.00 172
2018-02-19 17:48:16|> GPU7 57C Sol/s: 759.3 Sol/W: 2.50 Avg: 761.0 I/s: 406.9 Sh: 0.34 1.00 188 +
2018-02-19 17:48:19|> GPU8 55C Sol/s: 758.9 Sol/W: 2.76 Avg: 755.1 I/s: 404.9 Sh: 0.51 1.00 187
2018-02-19 17:48:21|> GPU9 62C Sol/s: 758.5 Sol/W: 2.56 Avg: 759.5 I/s: 406.1 Sh: 0.85 1.00 188
2018-02-19 17:48:21| ========== Sol/s: 7033.5 Sol/W: 2.65 Avg: 7056.1 I/s: 3778.3 Sh: 3.92 1.00 184

Hash will drop one one card in my rigs whenever I'm connected via an app remotely in Windows 10. Not sure if this is the same scenario you're seeing.

Sounds like it's pretty common? Maybe Remote Desktop uses the GPU itself as well. (see https://www.reddit.com/r/gpumining/comments/7eattb/mining_causes_poor_remote_desktop_teamviewer/)