Hmmm, is it normal that I have more hs/s with -t 2 -i 5.1 (640 threads x 2 = 1280) 650-680 than with single thread and -i 10.3 (1280 threads) - nearly 550 h/s ? I have single 1060 6 gb. In 1st case I look at summary of 2 threads [2017-12-10 04:04:28] accepted: 178/178 (diff 11.135), 649.20 H/s yes! [2017-12-10 04:04:45] GPU T0: Gigabyte GTX 1060 6GB, 471.93 H/s [2017-12-10 04:04:45] accepted: 179/179 (diff 10.517), 660.28 H/s yes! [2017-12-10 04:04:45] accepted: 180/180 (diff 6.498), 660.28 H/s yes! [2017-12-10 04:04:47] GPU T1: Gigabyte GTX 1060 6GB, 111.03 H/s [2017-12-10 04:04:47] accepted: 181/181 (diff 17.201), 654.67 H/s yes! [2017-12-10 04:04:55] GPU T1: Gigabyte GTX 1060 6GB, 414.54 H/s [2017-12-10 04:04:55] accepted: 182/182 (diff 9.587), 664.49 H/s yes! [2017-12-10 04:04:58] Stratum difficulty set to 8929 (8.929)
Am I missing smth? Also without any manual for intensity settings ccminer doesn't use all cuda cores of 1066 (1280 cores) and gives nearly 520 h/s.
|