Bitcoin Private - Pub Mining Pool


#1307

Nice, I’ll slowly work my way up those settings, ultimately I need the rig to be quiet as it’s in my livingroom next to my TV, I don’t want to upset my wife :wink: it already looks like a flashing Christmas tree.


#1308

Is that power draw including the motherboard etc? or is that total for the 6 GPU rig?


#1309

The overall power draw from the wall is 710-715W and includes the CPU and motherboard (USB stick, no HDD). The GPUs alone make-up 660W. I’m running a i3 Kabylake CPU (cheapest I could find) with 8Gb Ram.


#1310

impressive…i cant get below 900 with my 6 RX580 rig…according to my wattage meter


#1311

I’m currently running 880w on similar rig with solid 31.2 hash
https://ethermine.org/miners/03BEF224EAf3b41c18DAE95282a611879cf9c77d/dashboard


#1312

I just tried those settings and I’m getting average 3.8sols/w
Might be time to try smos and dstm

How many sols /card


#1313

Im getting around 490 Sols per Card


#1314

I must have an issue. I’m getting 435ish
Cheers :smiley:


#1315

I’ve got 12 gtx 1070 and Im at 1545 watts. my rig also has 10 extra fans on it too


#1316

gtx 1070. 470 sol/s, 4 sol/w so i’m not pounding on it bad but is not power hungry

core +150
mem +550
watts 115
455 to 470 sol/s


#1317

What os and miner do you use?
Cheers
Chris


#1318

smos on dtsm. gives the best shares draw back is smos is $2.00 per month.


#1319

OK, I’ve upper my settings to the following:
core=+175
mem=+750
power=115W
Hash= 4.3 Sols/W

Temperatures have increased a little, now sitting between 46-62C with fans at 45%.


#1320

Plus the dstm fee which can not be seen on the web view.

I have been experimenting with smos and nvoc and I have noticed a few things.

The temps with the same settings are much higher on smos and I had some stability issues.
The reported hash rates are different in the miners but when viewed on hashit they seem almost identical.
This is not scientific but the heat tells the story of actual power usage.


#1321

The actual power usage is the reading from the wall socket monitor. While the temperature is highly dependent on ambient room temperature (ranging from 20C to 35C), airflow and in my case sunlight from the window (makes a difference of 5-10C).


#1322

This was back to back just swapping out the usb drives a few times.
I should have plugged it into the meter though.


#1323

What temperature differences are you getting testing back-to-back?


#1324

On nvoc with the ti in the middle

59 68 65

Not sure of the exact figures on smos but the ti was on 76 (hitting the limiter) and the other two were within 2dgrees of 70.

The ventilation is not super good but it wasn’t an issue on nvoc


#1325

68C to 76C is quite a big difference. Have you checked the minimum fan setting and target temperature settings in both mining set-ups?

If you have different cards on either side of the 1070Ti in the middle you can modify the overclock settings individually. Just be careful if you swap the cards around on different slots afterwards, to avoid toasting them.


#1326

I might do some better testing later. Right now I’m letting the rig run so I can post a pic of the two different graphs in hashit.
I will leave all settings the same on all cards so there can be no confusion.

Probably not a fair caparison because we got a bunch of blocks during smos

CompareGraph

New graph. Colors have changed

Order is gpu 1,0,2 for o/c settings

Core 100,175,150
Mem 500,850,500
watt 120,115,115
temp targ 75
fan 45% min

gpu0 1070Ti
gpu1 1070
gpu2 1070

Miner Console
GPU2 59C Sol/s: 467.1 Sol/W: 4.03 Avg: 461.2 I/s: 246.9 Sh: 0.91 1.00 240
========== Sol/s: 1405.2 Sol/W: 4.02 Avg: 1401.3 I/s: 749.6 Sh: 2.78 1.00 239
gt; GPU0 64C Sol/s: 486.5 Sol/W: 4.19 Avg: 481.5 I/s: 257.6 Sh: 0.79 1.00 239
gt; GPU1 55C Sol/s: 447.0 Sol/W: 3.84 Avg: 458.4 I/s: 245.2 Sh: 1.05 1.00 239
gt; GPU2 59C Sol/s: 456.0 Sol/W: 4.03 Avg: 461.1 I/s: 246.9 Sh: 0.89 1.00 240
========== Sol/s: 1389.5 Sol/W: 4.02 Avg: 1401.1 I/s: 749.6 Sh: 2.73 1.00 239
GPU0 64C Sol/s: 482.1 Sol/W: 4.19 Avg: 481.5 I/s: 257.6 Sh: 0.83 1.00 238
GPU1 55C Sol/s: 459.1 Sol/W: 3.84 Avg: 458.5 I/s: 245.3 Sh: 1.03 1.00 239
GPU2 59C Sol/s: 466.8 Sol/W: 4.03 Avg: 461.2 I/s: 246.9 Sh: 0.93 1.00 240
========== Sol/s: 1408.0 Sol/W: 4.02 Avg: 1401.2 I/s: 749.8 Sh: 2.78 1.00 239
GPU0 64C Sol/s: 480.6 Sol/W: 4.19 Avg: 481.5 I/s: 257.6 Sh: 0.86 1.00 239
gt; GPU1 55C Sol/s: 458.5 Sol/W: 3.84 Avg: 458.5 I/s: 245.2 Sh: 1.06 1.00 239
gt; GPU2 59C Sol/s: 463.4 Sol/W: 4.03 Avg: 461.3 I/s: 246.9 Sh: 1.01 1.00 239
========== Sol/s: 1402.5 Sol/W: 4.02 Avg: 1401.2 I/s: 749.7 Sh: 2.94 1.00 239

Looks to me like the efficiency is exaggerated by smos
Plus you pay for smos
Plus you pay for dstm

I would like to see a long term test by someone with identical rigs but I will be staying with nvoc